rabbitmq-server/deps/rabbit/src/rabbit_fifo_client.erl

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1133 lines
48 KiB
Erlang
Raw Normal View History

%% This Source Code Form is subject to the terms of the Mozilla Public
%% License, v. 2.0. If a copy of the MPL was not distributed with this
%% file, You can obtain one at https://mozilla.org/MPL/2.0/.
%%
2024-01-02 11:02:20 +08:00
%% Copyright (c) 2007-2025 Broadcom. All Rights Reserved. The term “Broadcom” refers to Broadcom Inc. and/or its subsidiaries. All rights reserved.
%%
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Provides an easy to consume API for interacting with the {@link rabbit_fifo.}
%% state machine implementation running inside a `ra' raft system.
%%
%% Handles command tracking and other non-functional concerns.
-module(rabbit_fifo_client).
-include_lib("kernel/include/logger.hrl").
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
-export([
init/1,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
init/2,
close/1,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
checkout/4,
cancel_checkout/3,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
enqueue/3,
enqueue/4,
dequeue/4,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
settle/3,
return/3,
discard/3,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
modify/6,
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
credit_v1/4,
Introduce outbound RabbitMQ internal AMQP flow control ## What? Introduce RabbitMQ internal flow control for messages sent to AMQP clients. Prior this PR, when an AMQP client granted a large amount of link credit (e.g. 100k) to the sending queue, the sending queue sent that amount of messages to the session process no matter what. This becomes problematic for memory usage when the session process cannot send out messages fast enough to the AMQP client, especially if 1. The writer proc cannot send fast enough. This can happen when the AMQP client does not receive fast enough and causes TCP back-pressure to the server. Or 2. The server session proc is limited by remote-incoming-window. Both scenarios are now added as test cases. Tests * tcp_back_pressure_rabbitmq_internal_flow_quorum_queue * tcp_back_pressure_rabbitmq_internal_flow_classic_queue cover scenario 1. Tests * incoming_window_closed_rabbitmq_internal_flow_quorum_queue * incoming_window_closed_rabbitmq_internal_flow_classic_queue cover scenario 2. This PR sends messages from queues to AMQP clients in a more controlled manner. To illustrate: ``` make run-broker PLUGINS="rabbitmq_management" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 4" observer_cli:start() mq ``` where `mq` sorts by message queue length. Create a stream: ``` deps/rabbitmq_management/bin/rabbitmqadmin declare queue name=s1 queue_type=stream durable=true ``` Next, send and receive from the Stream via AMQP. Grant a large number of link credit to the sending stream: ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT bash-5.1# quiver //host.docker.internal//queue/s1 --durable -d 30s --credit 100000 ``` **Before** to this PR: ``` RESULTS Count ............................................... 100,696 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 120,422 messages/s Receiver rate ......................................... 3,363 messages/s End-to-end rate ....................................... 3,359 messages/s ``` We observe that all 100k link credit worth of messages are buffered in the writer proc's mailbox: ``` |No | Pid | MsgQueue |Name or Initial Call | Memory | Reductions |Current Function | |1 |<0.845.0> |100001 |rabbit_amqp_writer:init/1 | 126.0734 MB| 466633491 |prim_inet:send/5 | ``` **After** to this PR: ``` RESULTS Count ............................................. 2,973,440 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 123,322 messages/s Receiver rate ........................................ 99,250 messages/s End-to-end rate ...................................... 99,148 messages/s ``` We observe that the message queue lengths of both writer and session procs are low. ## How? Our goal is to have queues send out messages in a controlled manner without overloading RabbitMQ itself. We want RabbitMQ internal flow control between: ``` AMQP writer proc <--- session proc <--- queue proc ``` A similar concept exists for classic queues sending via AMQP 0.9.1. We want an approach that applies to AMQP and works generic for all queue types. For the interaction between AMQP writer proc and session proc we use a simple credit based approach reusing module `credit_flow`. For the interaction between session proc and queue proc, the following options exist: ### Option 1 The session process provides expliclity feedback to the queue after it has sent N messages. This approach is implemented in https://github.com/ansd/rabbitmq-server/tree/amqp-flow-control-poc-1 and works well. A new `rabbit_queue_type:sent/4` API was added which lets the queue proc know that it can send further messages to the session proc. Pros: * Will work equally well for AMQP 0.9.1, e.g. when quorum queues send messages in auto ack mode to AMQP 0.9.1 clients. * Simple for the session proc Cons: * Sligthly added complexity in every queue type implementation * Multiple Ra commands (settle, credit, sent) to decide when a quorum queue sends more messages. ### Option 2 A dual link approach where two AMQP links exists between ``` AMQP client <---link--> session proc <---link---> queue proc ``` When the client grants a large amount of credits, the session proc will top up credits to the queue proc periodically in smaller batches. Pros: * No queue type modifications required. * Re-uses AMQP link flow control Cons: * Significant added complexity in the session proc. A client can dynamically decrease or increase credits and dynamically change the drain mode while the session tops up credit to the queue. ### Option 3 Credit is a 32 bit unsigned integer. The spec mandates that the receiver independently chooses a credit. Nothing in the spec prevents the receiver to choose a credit of 1 billion. However the credit value is merely a **maximum**: > The link-credit variable defines the current maximum legal amount that the delivery-count can be increased by. Therefore, the server is not required to send all available messages to this receiver. For delivery-count: > Only the sender MAY independently modify this field. "independently" could be interpreted as the sender could add to the delivery-count irrespective of what the client chose for drain and link-credit. Option 3: The queue proc could at credit time already consume credit and advance the delivery-count if credit is too large before checking out any messages. For example if credit is 100k, but the queue only wants to send 1k, the queue could consume 99k of credits and advance the delivery-count, and subsequently send maximum 1k messages. If the queue advanced the delivery-count, RabbitMQ must send a FLOW to the receiver, otherwise the receiver wouldn’t know that it ran out of link-credit. Pros: * Very simple Cons: * Possibly unexpected behaviour for receiving AMQP clients * Possibly poor end-to-end throughput in auto-ack mode because the queue would send a batch of messages followed by a FLOW containing the advanced delivery-count. Only therafter the client will learn that it ran out of credits and top-up again. This feels like synchronously pulling a batch of messages. In contrast, option 2 sends out more messages as soon as the previous messages left RabbitMQ without requiring again a credit top up from the receiver. * drain mode with large credits requires the queue to send all available messages and only thereafter advance the delivery-count. Therefore, drain mode breaks option 3 somewhat. ### Option 4 Session proc drops message payload when its outgoing-pending queue gets too large and re-reads payloads from the queue once the message can be sent (see `get_checked_out` Ra command for quorum queues). Cons: * Would need to be implemented for every queue type, especially classic queues * Doesn't limit the amount of message metadata in the session proc's outgoing-pending queue ### Decision: Option 2 This commit implements option 2 to avoid any queue type modification. At most one credit request is in-flight between session process and queue process for a given queue consumer. If the AMQP client sends another FLOW in between, the session proc stashes the FLOW until it processes the previous credit reply. A delivery is only sent from the outgoing-pending queue if the session proc is not blocked by 1. writer proc, or 2. remote-incoming-window The credit reply is placed into the outgoing-pending queue. This ensures that the session proc will only top up the next batch of credits if sufficient messages were sent out to the writer proc. A future commit could additionally have each queue limit the number of unacked messages for a given AMQP consumer, or alternatively make use of session outgoing-window.
2024-05-18 17:11:12 +08:00
credit/5,
handle_ra_event/4,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
untracked_enqueue/2,
purge/1,
update_machine_state/2,
pending_size/1,
num_cached_segments/1,
stat/1,
stat/2,
Message Containers (#5077) This PR implements an approach for a "protocol (data format) agnostic core" where the format of the message isn't converted at point of reception. Currently all non AMQP 0.9.1 originating messages are converted into a AMQP 0.9.1 flavoured basic_message record before sent to a queue. If the messages are then consumed by the originating protocol they are converted back from AMQP 0.9.1. For some protocols such as MQTT 3.1 this isn't too expensive as MQTT is mostly a fairly easily mapped subset of AMQP 0.9.1 but for others such as AMQP 1.0 the conversions are awkward and in some cases lossy even if consuming from the originating protocol. This PR instead wraps all incoming messages in their originating form into a generic, extensible message container type (mc). The container module exposes an API to get common message details such as size and various properties (ttl, priority etc) directly from the source data type. Each protocol needs to implement the mc behaviour such that when a message originating form one protocol is consumed by another protocol we convert it to the target protocol at that point. The message container also contains annotations, dead letter records and other meta data we need to record during the lifetime of a message. The original protocol message is never modified unless it is consumed. This includes conversion modules to and from amqp, amqpl (AMQP 0.9.1) and mqtt. COMMIT HISTORY: * Refactor away from using the delivery{} record In many places including exchange types. This should make it easier to move towards using a message container type instead of basic_message. Add mc module and move direct replies outside of exchange Lots of changes incl classic queues Implement stream support incl amqp conversions simplify mc state record move mc.erl mc dlx stuff recent history exchange Make tracking work But doesn't take a protocol agnostic approach as we just convert everything into AMQP legacy and back. Might be good enough for now. Tracing as a whole may want a bit of a re-vamp at some point. tidy make quorum queue peek work by legacy conversion dead lettering fixes dead lettering fixes CMQ fixes rabbit_trace type fixes fixes fix Fix classic queue props test assertion fix feature flag and backwards compat Enable message_container feature flag in some SUITEs Dialyzer fixes fixes fix test fixes Various Manually update a gazelle generated file until a gazelle enhancement can be made https://github.com/rabbitmq/rules_erlang/issues/185 Add message_containers_SUITE to bazel and regen bazel files with gazelle from rules_erlang@main Simplify essential proprty access Such as durable, ttl and priority by extracting them into annotations at message container init time. Move type to remove dependenc on amqp10 stuff in mc.erl mostly because I don't know how to make bazel do the right thing add more stuff Refine routing header stuff wip Cosmetics Do not use "maybe" as type name as "maybe" is a keyword since OTP 25 which makes Erlang LS complain. * Dedup death queue names * Fix function clause crashes Fix failing tests in the MQTT shared_SUITE: A classic queue message ID can be undefined as set in https://github.com/rabbitmq/rabbitmq-server/blob/fbe79ff47b4edbc0fd95457e623d6593161ad198/deps/rabbit/src/rabbit_classic_queue_index_v2.erl#L1048 Fix failing tests in the MQTT shared_SUITE-mixed: When feature flag message_containers is disabled, the message is not an #mc{} record, but a #basic_message{} record. * Fix is_utf8_no_null crash Prior to this commit, the function crashed if invalid UTF-8 was provided, e.g.: ``` 1> rabbit_misc:is_valid_shortstr(<<"😇"/utf16>>). ** exception error: no function clause matching rabbit_misc:is_utf8_no_null(<<216,61,222,7>>) (rabbit_misc.erl, line 1481) ``` * Implement mqtt mc behaviour For now via amqp translation. This is still work in progress, but the following SUITEs pass: ``` make -C deps/rabbitmq_mqtt ct-shared t=[mqtt,v5,cluster_size_1] FULL=1 make -C deps/rabbitmq_mqtt ct-v5 t=[mqtt,cluster_size_1] FULL=1 ``` * Shorten mc file names Module name length matters because for each persistent message the #mc{} record is persisted to disk. ``` 1> iolist_size(term_to_iovec({mc, rabbit_mc_amqp_legacy})). 30 2> iolist_size(term_to_iovec({mc, mc_amqpl})). 17 ``` This commit renames the mc modules: ``` ag -l rabbit_mc_amqp_legacy | xargs sed -i 's/rabbit_mc_amqp_legacy/mc_amqpl/g' ag -l rabbit_mc_amqp | xargs sed -i 's/rabbit_mc_amqp/mc_amqp/g' ag -l rabbit_mqtt_mc | xargs sed -i 's/rabbit_mqtt_mc/mc_mqtt/g' ``` * mc: make deaths an annotation + fixes * Fix mc_mqtt protocol_state callback * Fix test will_delay_node_restart ``` make -C deps/rabbitmq_mqtt ct-v5 t=[mqtt,cluster_size_3]:will_delay_node_restart FULL=1 ``` * Bazel run gazelle * mix format rabbitmqctl.ex * Ensure ttl annotation is refelected in amqp legacy protocol state * Fix id access in message store * Fix rabbit_message_interceptor_SUITE * dializer fixes * Fix rabbit:rabbit_message_interceptor_SUITE-mixed set_annotation/3 should not result in duplicate keys * Fix MQTT shared_SUITE-mixed Up to 3.12 non-MQTT publishes were always QoS 1 regardless of delivery_mode. https://github.com/rabbitmq/rabbitmq-server/blob/75a953ce286a10aca910c098805a4f545989af38/deps/rabbitmq_mqtt/src/rabbit_mqtt_processor.erl#L2075-L2076 From now on, non-MQTT publishes are QoS 1 if durable. This makes more sense. The MQTT plugin must send a #basic_message{} to an old node that does not understand message containers. * Field content of 'v1_0.data' can be binary Fix ``` bazel test //deps/rabbitmq_mqtt:shared_SUITE-mixed \ --test_env FOCUS="-group [mqtt,v4,cluster_size_1] -case trace" \ -t- --test_sharding_strategy=disabled ``` * Remove route/2 and implement route/3 for all exchange types. This removes the route/2 callback from rabbit_exchange_type and makes route/3 mandatory instead. This is a breaking change and will require all implementations of exchange types to update their code, however this is necessary anyway for them to correctly handle the mc type. stream filtering fixes * Translate directly from MQTT to AMQP 0.9.1 * handle undecoded properties in mc_compat amqpl: put clause in right order recover death deatails from amqp data * Replace callback init_amqp with convert_from * Fix return value of lists:keyfind/3 * Translate directly from AMQP 0.9.1 to MQTT * Fix MQTT payload size MQTT payload can be a list when converted from AMQP 0.9.1 for example First conversions tests Plus some other conversion related fixes. bazel bazel translate amqp 1.0 null to undefined mc: property/2 and correlation_id/message_id return type tagged values. To ensure we can support a variety of types better. The type type tags are AMQP 1.0 flavoured. fix death recovery mc_mqtt: impl new api Add callbacks to allow protocols to compact data before storage And make readable if needing to query things repeatedly. bazel fix * more decoding * tracking mixed versions compat * mc: flip default of `durable` annotation to save some data. Assuming most messages are durable and that in memory messages suffer less from persistence overhead it makes sense for a non existent `durable` annotation to mean durable=true. * mc conversion tests and tidy up * mc make x_header unstrict again * amqpl: death record fixes * bazel * amqp -> amqpl conversion test * Fix crash in mc_amqp:size/1 Body can be a single amqp-value section (instead of being a list) as shown by test ``` make -C deps/rabbitmq_amqp1_0/ ct-system t=java ``` on branch native-amqp. * Fix crash in lists:flatten/1 Data can be a single amqp-value section (instead of being a list) as shown by test ``` make -C deps/rabbitmq_amqp1_0 ct-system t=dotnet:roundtrip_to_amqp_091 ``` on branch native-amqp. * Fix crash in rabbit_writer Running test ``` make -C deps/rabbitmq_amqp1_0 ct-system t=dotnet:roundtrip_to_amqp_091 ``` on branch native-amqp resulted in the following crash: ``` crasher: initial call: rabbit_writer:enter_mainloop/2 pid: <0.711.0> registered_name: [] exception error: bad argument in function size/1 called as size([<<0>>,<<"Sw">>,[<<160,2>>,<<"hi">>]]) *** argument 1: not tuple or binary in call from rabbit_binary_generator:build_content_frames/7 (rabbit_binary_generator.erl, line 89) in call from rabbit_binary_generator:build_simple_content_frames/4 (rabbit_binary_generator.erl, line 61) in call from rabbit_writer:assemble_frames/5 (rabbit_writer.erl, line 334) in call from rabbit_writer:internal_send_command_async/3 (rabbit_writer.erl, line 365) in call from rabbit_writer:handle_message/2 (rabbit_writer.erl, line 265) in call from rabbit_writer:handle_message/3 (rabbit_writer.erl, line 232) in call from rabbit_writer:mainloop1/2 (rabbit_writer.erl, line 223) ``` because #content.payload_fragments_rev is currently supposed to be a flat list of binaries instead of being an iolist. This commit fixes this crash inefficiently by calling iolist_to_binary/1. A better solution would be to allow AMQP legacy's #content.payload_fragments_rev to be an iolist. * Add accidentally deleted line back * mc: optimise mc_amqp internal format By removint the outer records for message and delivery annotations as well as application properties and footers. * mc: optimis mc_amqp map_add by using upsert * mc: refactoring and bug fixes * mc_SUITE routingheader assertions * mc remove serialize/1 callback as only used by amqp * mc_amqp: avoid returning a nested list from protocol_state * test and bug fix * move infer_type to mc_util * mc fixes and additiona assertions * Support headers exchange routing for MQTT messages When a headers exchange is bound to the MQTT topic exchange, routing will be performend based on both MQTT topic (by the topic exchange) and MQTT User Property (by the headers exchange). This combines the best worlds of both MQTT 5.0 and AMQP 0.9.1 and enables powerful routing topologies. When the User Property contains the same name multiple times, only the last name (and value) will be considered by the headers exchange. * Fix crash when sending from stream to amqpl When publishing a message via the stream protocol and consuming it via AMQP 0.9.1, the following crash occurred prior to this commit: ``` crasher: initial call: rabbit_channel:init/1 pid: <0.818.0> registered_name: [] exception exit: {{badmatch,undefined}, [{rabbit_channel,handle_deliver0,4, [{file,"rabbit_channel.erl"}, {line,2728}]}, {lists,foldl,3,[{file,"lists.erl"},{line,1594}]}, {rabbit_channel,handle_cast,2, [{file,"rabbit_channel.erl"}, {line,728}]}, {gen_server2,handle_msg,2, [{file,"gen_server2.erl"},{line,1056}]}, {proc_lib,wake_up,3, [{file,"proc_lib.erl"},{line,251}]}]} ``` This commit first gives `mc:init/3` the chance to set exchange and routing_keys annotations. If not set, `rabbit_stream_queue` will set these annotations assuming the message was originally published via the stream protocol. * Support consistent hash exchange routing for MQTT 5.0 When a consistent hash exchange is bound to the MQTT topic exchange, MQTT 5.0 messages can be routed to queues consistently based on the Correlation-Data in the PUBLISH packet. * Convert MQTT 5.0 User Property * to AMQP 0.9.1 headers * from AMQP 0.9.1 headers * to AMQP 1.0 application properties and message annotations * from AMQP 1.0 application properties and message annotations * Make use of Annotations in mc_mqtt:protocol_state/2 mc_mqtt:protocol_state/2 includes Annotations as parameter. It's cleaner to make use of these Annotations when computing the protocol state instead of relying on the caller (rabbitmq_mqtt_processor) to compute the protocol state. * Enforce AMQP 0.9.1 field name length limit The AMQP 0.9.1 spec prohibits field names longer than 128 characters. Therefore, when converting AMQP 1.0 message annotations, application properties or MQTT 5.0 User Property to AMQP 0.9.1 headers, drop any names longer than 128 characters. * Fix type specs Apply feedback from Michael Davis Co-authored-by: Michael Davis <mcarsondavis@gmail.com> * Add mc_mqtt unit test suite Implement mc_mqtt:x_header/2 * Translate indicator that payload is UTF-8 encoded when converting between MQTT 5.0 and AMQP 1.0 * Translate single amqp-value section from AMQP 1.0 to MQTT Convert to a text representation, if possible, and indicate to MQTT client that the payload is UTF-8 encoded. This way, the MQTT client will be able to parse the payload. If conversion to text representation is not possible, encode the payload using the AMQP 1.0 type system and indiate the encoding via Content-Type message/vnd.rabbitmq.amqp. This Content-Type is not registered. Type "message" makes sense since it's a message. Vendor tree "vnd.rabbitmq.amqp" makes sense since merely subtype "amqp" is not registered. * Fix payload conversion * Translate Response Topic between MQTT and AMQP Translate MQTT 5.0 Response Topic to AMQP 1.0 reply-to address and vice versa. The Response Topic must be a UTF-8 encoded string. This commit re-uses the already defined RabbitMQ target addresses: ``` "/topic/" RK Publish to amq.topic with routing key RK "/exchange/" X "/" RK Publish to exchange X with routing key RK ``` By default, the MQTT topic exchange is configure dto be amq.topic using the 1st target address. When an operator modifies the mqtt.exchange, the 2nd target address is used. * Apply PR feedback and fix formatting Co-authored-by: Michael Davis <mcarsondavis@gmail.com> * tidy up * Add MQTT message_containers test * consistent hash exchange: avoid amqp legacy conversion When hashing on a header value. * Avoid converting to amqp legacy when using exchange federation * Fix test flake * test and dialyzer fixes * dialyzer fix * Add MQTT protocol interoperability tests Test receiving from and sending to MQTT 5.0 and * AMQP 0.9.1 * AMQP 1.0 * STOMP * Streams * Regenerate portions of deps/rabbit/app.bzl with gazelle I'm not exactly sure how this happened, but gazell seems to have been run with an older version of the rules_erlang gazelle extension at some point. This caused generation of a structure that is no longer used. This commit updates the structure to the current pattern. * mc: refactoring * mc_amqpl: handle delivery annotations Just in case they are included. Also use iolist_to_iovec to create flat list of binaries when converting from amqp with amqp encoded payload. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com> Co-authored-by: Rin Kuryloski <kuryloskip@vmware.com>
2023-08-31 18:27:13 +08:00
query_single_active_consumer/1,
cluster_name/1
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
]).
-define(SOFT_LIMIT, 32).
-define(TIMER_TIME, 10000).
-define(COMMAND_TIMEOUT, 30000).
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
-define(UNLIMITED_PREFETCH_COUNT, 2000). %% something large for ra
%% controls the timer for closing cached segments
-define(CACHE_SEG_TIMEOUT, 5000).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
-type seq() :: non_neg_integer().
-type milliseconds() :: non_neg_integer().
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
-record(consumer, {key :: rabbit_fifo:consumer_key(),
% status = up :: up | cancelled,
last_msg_id :: seq() | -1 | undefined,
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
ack = false :: boolean(),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
%% Remove this field when feature flag rabbitmq_4.0.0 becomes required.
delivery_count :: {credit_api_v1, rabbit_queue_type:delivery_count()} |
credit_api_v2
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
}).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
-record(cfg, {servers = [] :: [ra:server_id()],
soft_limit = ?SOFT_LIMIT :: non_neg_integer(),
timeout :: non_neg_integer()
}).
-record(state, {cfg :: #cfg{},
leader :: undefined | ra:server_id(),
queue_status :: undefined | go | reject_publish,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
next_seq = 0 :: seq(),
next_enqueue_seq = 1 :: seq(),
%% indicates that we've exceeded the soft limit
slow = false :: boolean(),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
unsent_commands = #{} :: #{rabbit_fifo:consumer_key() =>
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
{[seq()], [seq()], [seq()]}},
pending = #{} :: #{seq() =>
{term(), rabbit_fifo:command()}},
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
consumers = #{} :: #{rabbit_types:ctag() => #consumer{}},
timer_state :: term(),
cached_segments :: undefined |
{undefined | reference(),
LastSeenMs :: milliseconds(),
ra_flru:state()}
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
}).
-opaque state() :: #state{}.
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-export_type([state/0]).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Create the initial state for a new rabbit_fifo sessions. A state is needed
%% to interact with a rabbit_fifo queue using @module.
%% @param Servers The known servers of the queue. If the current leader is known
%% ensure the leader node is at the head of the list.
-spec init([ra:server_id()]) -> state().
init(Servers) ->
init(Servers, ?SOFT_LIMIT).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Create the initial state for a new rabbit_fifo sessions. A state is needed
%% to interact with a rabbit_fifo queue using @module.
%% @param Servers The known servers of the queue. If the current leader is known
%% ensure the leader node is at the head of the list.
%% @param MaxPending size defining the max number of pending commands.
-spec init([ra:server_id()], non_neg_integer()) -> state().
init(Servers, SoftLimit) ->
2020-07-14 19:13:55 +08:00
Timeout = application:get_env(kernel, net_ticktime, 60) + 5,
#state{cfg = #cfg{servers = Servers,
soft_limit = SoftLimit,
2020-07-14 19:13:55 +08:00
timeout = Timeout * 1000}}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Enqueues a message.
%% @param QueueName Name of the queue.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @param Correlation an arbitrary erlang term used to correlate this
%% command when it has been applied.
%% @param Msg an arbitrary erlang term representing the message.
%% @param State the current {@module} state.
%% @returns
%% `{ok, State, Actions}' if the command was successfully sent
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% {@module} assigns a sequence number to every raft command it issues. The
%% SequenceNumber can be correlated to the applied sequence numbers returned
%% by the {@link handle_ra_event/2. handle_ra_event/2} function.
-spec enqueue(rabbit_amqqueue:name(), Correlation :: term(),
Msg :: term(), State :: state()) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{ok, state(), rabbit_queue_type:actions()} | {reject_publish, state()}.
enqueue(QName, Correlation, Msg,
#state{queue_status = undefined,
next_enqueue_seq = 1,
cfg = #cfg{servers = Servers,
timeout = Timeout}} = State0) ->
%% the first publish, register and enqueuer for this process.
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
%% TODO: we _only_ need to pre-register an enqueuer to discover if the
%% queue overflow is `reject_publish` and the queue can accept new messages
%% if the queue does not have `reject_publish` set we can skip this step
Reg = rabbit_fifo:make_register_enqueuer(self()),
case ra:process_command(Servers, Reg, Timeout) of
{ok, reject_publish, Leader} ->
{reject_publish, State0#state{leader = Leader,
queue_status = reject_publish}};
{ok, ok, Leader} ->
enqueue(QName, Correlation, Msg, State0#state{leader = Leader,
queue_status = go});
{error, {no_more_servers_to_try, _Errs}} ->
%% if we are not able to process the register command
%% it is safe to reject the message as we never attempted
%% to send it
{reject_publish, State0};
{error, {shutdown, delete}} ->
?LOG_DEBUG("~ts: QQ ~ts tried to register enqueuer during delete shutdown",
2025-07-16 06:32:20 +08:00
[?MODULE, rabbit_misc:rs(QName)]),
{reject_publish, State0};
{timeout, _} ->
{reject_publish, State0};
Err ->
?LOG_DEBUG("~ts: QQ ~ts error when registering enqueuer ~p",
2025-07-16 06:32:20 +08:00
[?MODULE, rabbit_misc:rs(QName), Err]),
exit(Err)
end;
enqueue(_QName, _Correlation, _Msg,
#state{queue_status = reject_publish,
cfg = #cfg{}} = State) ->
{reject_publish, State};
enqueue(QName, Correlation, Msg,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#state{slow = WasSlow,
pending = Pending,
2020-07-14 19:13:55 +08:00
queue_status = go,
next_seq = Seq,
next_enqueue_seq = EnqueueSeq,
cfg = #cfg{soft_limit = SftLmt}} = State0) ->
ServerId = pick_server(State0),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
% by default there is no correlation id
Cmd = rabbit_fifo:make_enqueue(self(), EnqueueSeq, Msg),
ok = ra:pipeline_command(ServerId, Cmd, Seq, low),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
IsSlow = map_size(Pending) >= SftLmt,
State = State0#state{pending = Pending#{Seq => {Correlation, Cmd}},
next_seq = Seq + 1,
next_enqueue_seq = EnqueueSeq + 1,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
slow = IsSlow},
if IsSlow andalso not WasSlow ->
{ok, set_timer(QName, State), [{block, cluster_name(State)}]};
true ->
{ok, State, []}
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
end.
%% @doc Enqueues a message.
%% @param QueueName Name of the queue.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @param Msg an arbitrary erlang term representing the message.
%% @param State the current {@module} state.
%% @return's
%% `{ok, State, Actions}' if the command was successfully sent.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% {@module} assigns a sequence number to every raft command it issues. The
%% SequenceNumber can be correlated to the applied sequence numbers returned
%% by the {@link handle_ra_event/2. handle_ra_event/2} function.
%%
-spec enqueue(rabbit_amqqueue:name(), Msg :: term(), State :: state()) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{ok, state(), rabbit_queue_type:actions()} | {reject_publish, state()}.
enqueue(QName, Msg, State) ->
enqueue(QName, undefined, Msg, State).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Dequeue a message from the queue.
%%
2019-02-13 01:28:22 +08:00
%% This is a synchronous call. I.e. the call will block until the command
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% has been accepted by the ra process or it times out.
%%
%% @param QueueName Name of the queue.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @param ConsumerTag a unique tag to identify this particular consumer.
%% @param Settlement either `settled' or `unsettled'. When `settled' no
%% further settlement needs to be done.
%% @param State The {@module} state.
%%
%% @returns `{ok, IdMsg, State}' or `{error | timeout, term()}'
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec dequeue(rabbit_amqqueue:name(), rabbit_types:ctag(),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Settlement :: settled | unsettled, state()) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{ok, non_neg_integer(), term(), non_neg_integer()}
| {empty, state()} | {error | timeout, term()}.
dequeue(QueueName, ConsumerTag, Settlement,
#state{cfg = #cfg{timeout = Timeout}} = State0) ->
ServerId = pick_server(State0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
%% dequeue never really needs to assign a consumer key so we just use
%% the old ConsumerId format here
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
ConsumerId = consumer_id(ConsumerTag),
case ra:process_command(ServerId,
rabbit_fifo:make_checkout(ConsumerId,
{dequeue, Settlement},
#{}),
Timeout) of
{ok, {dequeue, empty}, Leader} ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{empty, State0#state{leader = Leader}};
{ok, {dequeue, {MsgId, {MsgHeader, Msg0}}, MsgsReady}, Leader} ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
{Msg, Redelivered} = add_delivery_count_header(Msg0, MsgHeader),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{ok, MsgsReady,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
{QueueName, qref(Leader), MsgId, Redelivered, Msg},
State0#state{leader = Leader}};
{ok, {error, _} = Err, _Leader} ->
Err;
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Err ->
Err
end.
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
add_delivery_count_header(Msg0, #{acquired_count := AcqCount} = Header)
when is_integer(AcqCount) ->
Msg = case mc:is(Msg0) of
true ->
Msg1 = mc:set_annotation(<<"x-delivery-count">>, AcqCount, Msg0),
%% the "delivery-count" header in the AMQP spec does not include
%% returns (released outcomes)
rabbit_fifo:annotate_msg(Header, Msg1);
false ->
Msg0
end,
Redelivered = AcqCount > 0,
{Msg, Redelivered};
add_delivery_count_header(Msg, #{delivery_count := DC} = Header) ->
%% there was a delivery count but no acquired count, this means the message
%% was delivered from a quorum queue running v3 so we patch this up here
add_delivery_count_header(Msg, Header#{acquired_count => DC});
add_delivery_count_header(Msg, _Header) ->
{Msg, false}.
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Settle a message. Permanently removes message from the queue.
%% @param ConsumerTag the tag uniquely identifying the consumer.
%% @param MsgIds the message ids received with the {@link rabbit_fifo:delivery/0.}
%% @param State the {@module} state
%%
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec settle(rabbit_types:ctag(), [rabbit_fifo:msg_id()], state()) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{state(), list()}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
settle(ConsumerTag, [_|_] = MsgIds, #state{slow = false} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
ServerId = pick_server(State0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Cmd = rabbit_fifo:make_settle(ConsumerKey, MsgIds),
{send_command(ServerId, undefined, Cmd, normal, State0), []};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
settle(ConsumerTag, [_|_] = MsgIds,
#state{unsent_commands = Unsent0} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% we've reached the soft limit so will stash the command to be
%% sent once we have seen enough notifications
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Unsent = maps:update_with(ConsumerKey,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
fun ({Settles, Returns, Discards}) ->
%% MsgIds has fewer elements than Settles.
%% Therefore put it on the left side of the ++ operator.
%% The order in which messages are settled does not matter.
{MsgIds ++ Settles, Returns, Discards}
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
end, {MsgIds, [], []}, Unsent0),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{State0#state{unsent_commands = Unsent}, []}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Return a message to the queue.
%% @param ConsumerTag the tag uniquely identifying the consumer.
%% @param MsgIds the message ids to return received
%% from {@link rabbit_fifo:delivery/0.}
%% @param State the {@module} state
%% @returns
%% `{State, list()}' if the command was successfully sent.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%%
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec return(rabbit_types:ctag(), [rabbit_fifo:msg_id()], state()) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{state(), list()}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
return(ConsumerTag, [_|_] = MsgIds, #state{slow = false} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
ServerId = pick_server(State0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Cmd = rabbit_fifo:make_return(ConsumerKey, MsgIds),
{send_command(ServerId, undefined, Cmd, normal, State0), []};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
return(ConsumerTag, [_|_] = MsgIds,
#state{unsent_commands = Unsent0} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% we've reached the soft limit so will stash the command to be
%% sent once we have seen enough notifications
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Unsent = maps:update_with(ConsumerKey,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
fun ({Settles, Returns, Discards}) ->
{Settles, Returns ++ MsgIds, Discards}
end, {[], MsgIds, []}, Unsent0),
State1 = State0#state{unsent_commands = Unsent},
{State1, []}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Discards a checked out message.
%% If the queue has a dead_letter_handler configured this will be called.
%% @param ConsumerTag the tag uniquely identifying the consumer.
%% @param MsgIds the message ids to discard
%% from {@link rabbit_fifo:delivery/0.}
%% @param State the {@module} state
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec discard(rabbit_types:ctag(), [rabbit_fifo:msg_id()], state()) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{state(), list()}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
discard(ConsumerTag, [_|_] = MsgIds, #state{slow = false} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
ServerId = pick_server(State0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Cmd = rabbit_fifo:make_discard(ConsumerKey, MsgIds),
{send_command(ServerId, undefined, Cmd, normal, State0), []};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
discard(ConsumerTag, [_|_] = MsgIds,
#state{unsent_commands = Unsent0} = State0) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State0),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% we've reached the soft limit so will stash the command to be
%% sent once we have seen enough notifications
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Unsent = maps:update_with(ConsumerKey,
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
fun ({Settles, Returns, Discards}) ->
{Settles, Returns, Discards ++ MsgIds}
end, {[], [], MsgIds}, Unsent0),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{State0#state{unsent_commands = Unsent}, []}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
-spec modify(rabbit_types:ctag(), [rabbit_fifo:msg_id()],
boolean(), boolean(), mc:annotations(), state()) ->
{state(), list()}.
modify(ConsumerTag, [_|_] = MsgIds, DelFailed, Undel, Anns,
#state{} = State0) ->
ConsumerKey = consumer_key(ConsumerTag, State0),
%% we need to send any pending settles, discards or returns before we
%% send the modify as this cannot be batched
%% as it contains message specific annotations
State1 = send_pending(ConsumerKey, State0),
ServerId = pick_server(State1),
Cmd = rabbit_fifo:make_modify(ConsumerKey, MsgIds, DelFailed, Undel, Anns),
{send_command(ServerId, undefined, Cmd, normal, State1), []}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Register with the rabbit_fifo queue to "checkout" messages as they
%% become available.
%%
2019-02-13 01:28:22 +08:00
%% This is a synchronous call. I.e. the call will block until the command
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% has been accepted by the ra process or it times out.
%%
%% @param ConsumerTag a unique tag to identify this particular consumer.
%% @param NumUnsettled the maximum number of in-flight messages. Once this
%% number of messages has been received but not settled no further messages
%% will be delivered to the consumer.
%% @param CreditMode The credit mode to use for the checkout.
%% simple_prefetch: credit is auto topped up as deliveries are settled
%% credited: credit is only increased by sending credit to the queue
%% @param State The {@module} state.
%%
%% @returns `{ok, State}' or `{error | timeout, term()}'
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec checkout(rabbit_types:ctag(),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
CreditMode :: rabbit_fifo:credit_mode(),
Meta :: rabbit_fifo:consumer_meta(),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
state()) ->
{ok, ConsumerInfos :: map(), state()} |
{error | timeout, term()}.
checkout(ConsumerTag, CreditMode, #{} = Meta,
#state{consumers = CDels0} = State0)
when is_binary(ConsumerTag) andalso
is_tuple(CreditMode) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Servers = sorted_servers(State0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerId = consumer_id(ConsumerTag),
Spec = case rabbit_fifo:is_v4() of
true ->
case CreditMode of
{simple_prefetch, 0} ->
{auto, {simple_prefetch,
?UNLIMITED_PREFETCH_COUNT}};
_ ->
{auto, CreditMode}
end;
false ->
case CreditMode of
{credited, _} ->
{auto, 0, credited};
{simple_prefetch, 0} ->
{auto, ?UNLIMITED_PREFETCH_COUNT, simple_prefetch};
{simple_prefetch, Num} ->
{auto, Num, simple_prefetch}
end
end,
Cmd = rabbit_fifo:make_checkout(ConsumerId, Spec, Meta),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
%% ???
Ack = maps:get(ack, Meta, true),
case try_process_command(Servers, Cmd, State0) of
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
{ok, {ok, Reply}, Leader} ->
LastMsgId = case Reply of
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#{num_checked_out := NumChecked,
next_msg_id := NextMsgId} ->
case NumChecked > 0 of
true ->
%% we cannot know if the pending messages
%% have been delivered to the client or they
%% are on their way to the current process.
%% We set `undefined' to signal this uncertainty
%% and will just accept the next arriving message
%% irrespective of message id
undefined;
false ->
NextMsgId - 1
end
end,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
DeliveryCount = case rabbit_fifo:is_v4() of
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
true -> credit_api_v2;
false -> {credit_api_v1, 0}
end,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = maps:get(key, Reply, ConsumerId),
SDels = maps:update_with(
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
ConsumerTag,
fun (C) -> C#consumer{ack = Ack} end,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#consumer{key = ConsumerKey,
last_msg_id = LastMsgId,
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
ack = Ack,
delivery_count = DeliveryCount},
CDels0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
{ok, Reply, State0#state{leader = Leader,
consumers = SDels}};
Err ->
Err
end.
-spec query_single_active_consumer(state()) ->
{ok, term()} | {error, term()} | {timeout, term()}.
query_single_active_consumer(#state{leader = undefined}) ->
{error, leader_not_known};
query_single_active_consumer(#state{leader = Leader}) ->
case ra:local_query(Leader, fun rabbit_fifo:query_single_active_consumer/1,
?COMMAND_TIMEOUT) of
{ok, {_, Reply}, _} ->
{ok, Reply};
Err ->
Err
end.
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
-spec credit_v1(rabbit_types:ctag(),
Credit :: non_neg_integer(),
Drain :: boolean(),
state()) ->
{state(), rabbit_queue_type:actions()}.
credit_v1(ConsumerTag, Credit, Drain,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#state{consumers = CDels} = State) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
#consumer{delivery_count = {credit_api_v1, Count}} = maps:get(ConsumerTag, CDels),
Introduce outbound RabbitMQ internal AMQP flow control ## What? Introduce RabbitMQ internal flow control for messages sent to AMQP clients. Prior this PR, when an AMQP client granted a large amount of link credit (e.g. 100k) to the sending queue, the sending queue sent that amount of messages to the session process no matter what. This becomes problematic for memory usage when the session process cannot send out messages fast enough to the AMQP client, especially if 1. The writer proc cannot send fast enough. This can happen when the AMQP client does not receive fast enough and causes TCP back-pressure to the server. Or 2. The server session proc is limited by remote-incoming-window. Both scenarios are now added as test cases. Tests * tcp_back_pressure_rabbitmq_internal_flow_quorum_queue * tcp_back_pressure_rabbitmq_internal_flow_classic_queue cover scenario 1. Tests * incoming_window_closed_rabbitmq_internal_flow_quorum_queue * incoming_window_closed_rabbitmq_internal_flow_classic_queue cover scenario 2. This PR sends messages from queues to AMQP clients in a more controlled manner. To illustrate: ``` make run-broker PLUGINS="rabbitmq_management" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 4" observer_cli:start() mq ``` where `mq` sorts by message queue length. Create a stream: ``` deps/rabbitmq_management/bin/rabbitmqadmin declare queue name=s1 queue_type=stream durable=true ``` Next, send and receive from the Stream via AMQP. Grant a large number of link credit to the sending stream: ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT bash-5.1# quiver //host.docker.internal//queue/s1 --durable -d 30s --credit 100000 ``` **Before** to this PR: ``` RESULTS Count ............................................... 100,696 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 120,422 messages/s Receiver rate ......................................... 3,363 messages/s End-to-end rate ....................................... 3,359 messages/s ``` We observe that all 100k link credit worth of messages are buffered in the writer proc's mailbox: ``` |No | Pid | MsgQueue |Name or Initial Call | Memory | Reductions |Current Function | |1 |<0.845.0> |100001 |rabbit_amqp_writer:init/1 | 126.0734 MB| 466633491 |prim_inet:send/5 | ``` **After** to this PR: ``` RESULTS Count ............................................. 2,973,440 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 123,322 messages/s Receiver rate ........................................ 99,250 messages/s End-to-end rate ...................................... 99,148 messages/s ``` We observe that the message queue lengths of both writer and session procs are low. ## How? Our goal is to have queues send out messages in a controlled manner without overloading RabbitMQ itself. We want RabbitMQ internal flow control between: ``` AMQP writer proc <--- session proc <--- queue proc ``` A similar concept exists for classic queues sending via AMQP 0.9.1. We want an approach that applies to AMQP and works generic for all queue types. For the interaction between AMQP writer proc and session proc we use a simple credit based approach reusing module `credit_flow`. For the interaction between session proc and queue proc, the following options exist: ### Option 1 The session process provides expliclity feedback to the queue after it has sent N messages. This approach is implemented in https://github.com/ansd/rabbitmq-server/tree/amqp-flow-control-poc-1 and works well. A new `rabbit_queue_type:sent/4` API was added which lets the queue proc know that it can send further messages to the session proc. Pros: * Will work equally well for AMQP 0.9.1, e.g. when quorum queues send messages in auto ack mode to AMQP 0.9.1 clients. * Simple for the session proc Cons: * Sligthly added complexity in every queue type implementation * Multiple Ra commands (settle, credit, sent) to decide when a quorum queue sends more messages. ### Option 2 A dual link approach where two AMQP links exists between ``` AMQP client <---link--> session proc <---link---> queue proc ``` When the client grants a large amount of credits, the session proc will top up credits to the queue proc periodically in smaller batches. Pros: * No queue type modifications required. * Re-uses AMQP link flow control Cons: * Significant added complexity in the session proc. A client can dynamically decrease or increase credits and dynamically change the drain mode while the session tops up credit to the queue. ### Option 3 Credit is a 32 bit unsigned integer. The spec mandates that the receiver independently chooses a credit. Nothing in the spec prevents the receiver to choose a credit of 1 billion. However the credit value is merely a **maximum**: > The link-credit variable defines the current maximum legal amount that the delivery-count can be increased by. Therefore, the server is not required to send all available messages to this receiver. For delivery-count: > Only the sender MAY independently modify this field. "independently" could be interpreted as the sender could add to the delivery-count irrespective of what the client chose for drain and link-credit. Option 3: The queue proc could at credit time already consume credit and advance the delivery-count if credit is too large before checking out any messages. For example if credit is 100k, but the queue only wants to send 1k, the queue could consume 99k of credits and advance the delivery-count, and subsequently send maximum 1k messages. If the queue advanced the delivery-count, RabbitMQ must send a FLOW to the receiver, otherwise the receiver wouldn’t know that it ran out of link-credit. Pros: * Very simple Cons: * Possibly unexpected behaviour for receiving AMQP clients * Possibly poor end-to-end throughput in auto-ack mode because the queue would send a batch of messages followed by a FLOW containing the advanced delivery-count. Only therafter the client will learn that it ran out of credits and top-up again. This feels like synchronously pulling a batch of messages. In contrast, option 2 sends out more messages as soon as the previous messages left RabbitMQ without requiring again a credit top up from the receiver. * drain mode with large credits requires the queue to send all available messages and only thereafter advance the delivery-count. Therefore, drain mode breaks option 3 somewhat. ### Option 4 Session proc drops message payload when its outgoing-pending queue gets too large and re-reads payloads from the queue once the message can be sent (see `get_checked_out` Ra command for quorum queues). Cons: * Would need to be implemented for every queue type, especially classic queues * Doesn't limit the amount of message metadata in the session proc's outgoing-pending queue ### Decision: Option 2 This commit implements option 2 to avoid any queue type modification. At most one credit request is in-flight between session process and queue process for a given queue consumer. If the AMQP client sends another FLOW in between, the session proc stashes the FLOW until it processes the previous credit reply. A delivery is only sent from the outgoing-pending queue if the session proc is not blocked by 1. writer proc, or 2. remote-incoming-window The credit reply is placed into the outgoing-pending queue. This ensures that the session proc will only top up the next batch of credits if sufficient messages were sent out to the writer proc. A future commit could additionally have each queue limit the number of unacked messages for a given AMQP consumer, or alternatively make use of session outgoing-window.
2024-05-18 17:11:12 +08:00
credit(ConsumerTag, Count, Credit, Drain, State).
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Provide credit to the queue
%%
%% This only has an effect if the consumer uses credit mode: credited
%% @param ConsumerTag a unique tag to identify this particular consumer.
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
%% @param Credit the amount of credit to provide to the queue
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @param Drain tells the queue to use up any credit that cannot be immediately
%% fulfilled. (i.e. there are not enough messages on queue to use up all the
%% provided credit).
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
%% @param Reply true if the queue client requests a credit_reply queue action
-spec credit(rabbit_types:ctag(),
rabbit_queue_type:delivery_count(),
rabbit_queue_type:credit(),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Drain :: boolean(),
state()) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{state(), rabbit_queue_type:actions()}.
Introduce outbound RabbitMQ internal AMQP flow control ## What? Introduce RabbitMQ internal flow control for messages sent to AMQP clients. Prior this PR, when an AMQP client granted a large amount of link credit (e.g. 100k) to the sending queue, the sending queue sent that amount of messages to the session process no matter what. This becomes problematic for memory usage when the session process cannot send out messages fast enough to the AMQP client, especially if 1. The writer proc cannot send fast enough. This can happen when the AMQP client does not receive fast enough and causes TCP back-pressure to the server. Or 2. The server session proc is limited by remote-incoming-window. Both scenarios are now added as test cases. Tests * tcp_back_pressure_rabbitmq_internal_flow_quorum_queue * tcp_back_pressure_rabbitmq_internal_flow_classic_queue cover scenario 1. Tests * incoming_window_closed_rabbitmq_internal_flow_quorum_queue * incoming_window_closed_rabbitmq_internal_flow_classic_queue cover scenario 2. This PR sends messages from queues to AMQP clients in a more controlled manner. To illustrate: ``` make run-broker PLUGINS="rabbitmq_management" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 4" observer_cli:start() mq ``` where `mq` sorts by message queue length. Create a stream: ``` deps/rabbitmq_management/bin/rabbitmqadmin declare queue name=s1 queue_type=stream durable=true ``` Next, send and receive from the Stream via AMQP. Grant a large number of link credit to the sending stream: ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT bash-5.1# quiver //host.docker.internal//queue/s1 --durable -d 30s --credit 100000 ``` **Before** to this PR: ``` RESULTS Count ............................................... 100,696 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 120,422 messages/s Receiver rate ......................................... 3,363 messages/s End-to-end rate ....................................... 3,359 messages/s ``` We observe that all 100k link credit worth of messages are buffered in the writer proc's mailbox: ``` |No | Pid | MsgQueue |Name or Initial Call | Memory | Reductions |Current Function | |1 |<0.845.0> |100001 |rabbit_amqp_writer:init/1 | 126.0734 MB| 466633491 |prim_inet:send/5 | ``` **After** to this PR: ``` RESULTS Count ............................................. 2,973,440 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 123,322 messages/s Receiver rate ........................................ 99,250 messages/s End-to-end rate ...................................... 99,148 messages/s ``` We observe that the message queue lengths of both writer and session procs are low. ## How? Our goal is to have queues send out messages in a controlled manner without overloading RabbitMQ itself. We want RabbitMQ internal flow control between: ``` AMQP writer proc <--- session proc <--- queue proc ``` A similar concept exists for classic queues sending via AMQP 0.9.1. We want an approach that applies to AMQP and works generic for all queue types. For the interaction between AMQP writer proc and session proc we use a simple credit based approach reusing module `credit_flow`. For the interaction between session proc and queue proc, the following options exist: ### Option 1 The session process provides expliclity feedback to the queue after it has sent N messages. This approach is implemented in https://github.com/ansd/rabbitmq-server/tree/amqp-flow-control-poc-1 and works well. A new `rabbit_queue_type:sent/4` API was added which lets the queue proc know that it can send further messages to the session proc. Pros: * Will work equally well for AMQP 0.9.1, e.g. when quorum queues send messages in auto ack mode to AMQP 0.9.1 clients. * Simple for the session proc Cons: * Sligthly added complexity in every queue type implementation * Multiple Ra commands (settle, credit, sent) to decide when a quorum queue sends more messages. ### Option 2 A dual link approach where two AMQP links exists between ``` AMQP client <---link--> session proc <---link---> queue proc ``` When the client grants a large amount of credits, the session proc will top up credits to the queue proc periodically in smaller batches. Pros: * No queue type modifications required. * Re-uses AMQP link flow control Cons: * Significant added complexity in the session proc. A client can dynamically decrease or increase credits and dynamically change the drain mode while the session tops up credit to the queue. ### Option 3 Credit is a 32 bit unsigned integer. The spec mandates that the receiver independently chooses a credit. Nothing in the spec prevents the receiver to choose a credit of 1 billion. However the credit value is merely a **maximum**: > The link-credit variable defines the current maximum legal amount that the delivery-count can be increased by. Therefore, the server is not required to send all available messages to this receiver. For delivery-count: > Only the sender MAY independently modify this field. "independently" could be interpreted as the sender could add to the delivery-count irrespective of what the client chose for drain and link-credit. Option 3: The queue proc could at credit time already consume credit and advance the delivery-count if credit is too large before checking out any messages. For example if credit is 100k, but the queue only wants to send 1k, the queue could consume 99k of credits and advance the delivery-count, and subsequently send maximum 1k messages. If the queue advanced the delivery-count, RabbitMQ must send a FLOW to the receiver, otherwise the receiver wouldn’t know that it ran out of link-credit. Pros: * Very simple Cons: * Possibly unexpected behaviour for receiving AMQP clients * Possibly poor end-to-end throughput in auto-ack mode because the queue would send a batch of messages followed by a FLOW containing the advanced delivery-count. Only therafter the client will learn that it ran out of credits and top-up again. This feels like synchronously pulling a batch of messages. In contrast, option 2 sends out more messages as soon as the previous messages left RabbitMQ without requiring again a credit top up from the receiver. * drain mode with large credits requires the queue to send all available messages and only thereafter advance the delivery-count. Therefore, drain mode breaks option 3 somewhat. ### Option 4 Session proc drops message payload when its outgoing-pending queue gets too large and re-reads payloads from the queue once the message can be sent (see `get_checked_out` Ra command for quorum queues). Cons: * Would need to be implemented for every queue type, especially classic queues * Doesn't limit the amount of message metadata in the session proc's outgoing-pending queue ### Decision: Option 2 This commit implements option 2 to avoid any queue type modification. At most one credit request is in-flight between session process and queue process for a given queue consumer. If the AMQP client sends another FLOW in between, the session proc stashes the FLOW until it processes the previous credit reply. A delivery is only sent from the outgoing-pending queue if the session proc is not blocked by 1. writer proc, or 2. remote-incoming-window The credit reply is placed into the outgoing-pending queue. This ensures that the session proc will only top up the next batch of credits if sufficient messages were sent out to the writer proc. A future commit could additionally have each queue limit the number of unacked messages for a given AMQP consumer, or alternatively make use of session outgoing-window.
2024-05-18 17:11:12 +08:00
credit(ConsumerTag, DeliveryCount, Credit, Drain, State) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State),
Introduce outbound RabbitMQ internal AMQP flow control ## What? Introduce RabbitMQ internal flow control for messages sent to AMQP clients. Prior this PR, when an AMQP client granted a large amount of link credit (e.g. 100k) to the sending queue, the sending queue sent that amount of messages to the session process no matter what. This becomes problematic for memory usage when the session process cannot send out messages fast enough to the AMQP client, especially if 1. The writer proc cannot send fast enough. This can happen when the AMQP client does not receive fast enough and causes TCP back-pressure to the server. Or 2. The server session proc is limited by remote-incoming-window. Both scenarios are now added as test cases. Tests * tcp_back_pressure_rabbitmq_internal_flow_quorum_queue * tcp_back_pressure_rabbitmq_internal_flow_classic_queue cover scenario 1. Tests * incoming_window_closed_rabbitmq_internal_flow_quorum_queue * incoming_window_closed_rabbitmq_internal_flow_classic_queue cover scenario 2. This PR sends messages from queues to AMQP clients in a more controlled manner. To illustrate: ``` make run-broker PLUGINS="rabbitmq_management" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 4" observer_cli:start() mq ``` where `mq` sorts by message queue length. Create a stream: ``` deps/rabbitmq_management/bin/rabbitmqadmin declare queue name=s1 queue_type=stream durable=true ``` Next, send and receive from the Stream via AMQP. Grant a large number of link credit to the sending stream: ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT bash-5.1# quiver //host.docker.internal//queue/s1 --durable -d 30s --credit 100000 ``` **Before** to this PR: ``` RESULTS Count ............................................... 100,696 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 120,422 messages/s Receiver rate ......................................... 3,363 messages/s End-to-end rate ....................................... 3,359 messages/s ``` We observe that all 100k link credit worth of messages are buffered in the writer proc's mailbox: ``` |No | Pid | MsgQueue |Name or Initial Call | Memory | Reductions |Current Function | |1 |<0.845.0> |100001 |rabbit_amqp_writer:init/1 | 126.0734 MB| 466633491 |prim_inet:send/5 | ``` **After** to this PR: ``` RESULTS Count ............................................. 2,973,440 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 123,322 messages/s Receiver rate ........................................ 99,250 messages/s End-to-end rate ...................................... 99,148 messages/s ``` We observe that the message queue lengths of both writer and session procs are low. ## How? Our goal is to have queues send out messages in a controlled manner without overloading RabbitMQ itself. We want RabbitMQ internal flow control between: ``` AMQP writer proc <--- session proc <--- queue proc ``` A similar concept exists for classic queues sending via AMQP 0.9.1. We want an approach that applies to AMQP and works generic for all queue types. For the interaction between AMQP writer proc and session proc we use a simple credit based approach reusing module `credit_flow`. For the interaction between session proc and queue proc, the following options exist: ### Option 1 The session process provides expliclity feedback to the queue after it has sent N messages. This approach is implemented in https://github.com/ansd/rabbitmq-server/tree/amqp-flow-control-poc-1 and works well. A new `rabbit_queue_type:sent/4` API was added which lets the queue proc know that it can send further messages to the session proc. Pros: * Will work equally well for AMQP 0.9.1, e.g. when quorum queues send messages in auto ack mode to AMQP 0.9.1 clients. * Simple for the session proc Cons: * Sligthly added complexity in every queue type implementation * Multiple Ra commands (settle, credit, sent) to decide when a quorum queue sends more messages. ### Option 2 A dual link approach where two AMQP links exists between ``` AMQP client <---link--> session proc <---link---> queue proc ``` When the client grants a large amount of credits, the session proc will top up credits to the queue proc periodically in smaller batches. Pros: * No queue type modifications required. * Re-uses AMQP link flow control Cons: * Significant added complexity in the session proc. A client can dynamically decrease or increase credits and dynamically change the drain mode while the session tops up credit to the queue. ### Option 3 Credit is a 32 bit unsigned integer. The spec mandates that the receiver independently chooses a credit. Nothing in the spec prevents the receiver to choose a credit of 1 billion. However the credit value is merely a **maximum**: > The link-credit variable defines the current maximum legal amount that the delivery-count can be increased by. Therefore, the server is not required to send all available messages to this receiver. For delivery-count: > Only the sender MAY independently modify this field. "independently" could be interpreted as the sender could add to the delivery-count irrespective of what the client chose for drain and link-credit. Option 3: The queue proc could at credit time already consume credit and advance the delivery-count if credit is too large before checking out any messages. For example if credit is 100k, but the queue only wants to send 1k, the queue could consume 99k of credits and advance the delivery-count, and subsequently send maximum 1k messages. If the queue advanced the delivery-count, RabbitMQ must send a FLOW to the receiver, otherwise the receiver wouldn’t know that it ran out of link-credit. Pros: * Very simple Cons: * Possibly unexpected behaviour for receiving AMQP clients * Possibly poor end-to-end throughput in auto-ack mode because the queue would send a batch of messages followed by a FLOW containing the advanced delivery-count. Only therafter the client will learn that it ran out of credits and top-up again. This feels like synchronously pulling a batch of messages. In contrast, option 2 sends out more messages as soon as the previous messages left RabbitMQ without requiring again a credit top up from the receiver. * drain mode with large credits requires the queue to send all available messages and only thereafter advance the delivery-count. Therefore, drain mode breaks option 3 somewhat. ### Option 4 Session proc drops message payload when its outgoing-pending queue gets too large and re-reads payloads from the queue once the message can be sent (see `get_checked_out` Ra command for quorum queues). Cons: * Would need to be implemented for every queue type, especially classic queues * Doesn't limit the amount of message metadata in the session proc's outgoing-pending queue ### Decision: Option 2 This commit implements option 2 to avoid any queue type modification. At most one credit request is in-flight between session process and queue process for a given queue consumer. If the AMQP client sends another FLOW in between, the session proc stashes the FLOW until it processes the previous credit reply. A delivery is only sent from the outgoing-pending queue if the session proc is not blocked by 1. writer proc, or 2. remote-incoming-window The credit reply is placed into the outgoing-pending queue. This ensures that the session proc will only top up the next batch of credits if sufficient messages were sent out to the writer proc. A future commit could additionally have each queue limit the number of unacked messages for a given AMQP consumer, or alternatively make use of session outgoing-window.
2024-05-18 17:11:12 +08:00
ServerId = pick_server(State),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Cmd = rabbit_fifo:make_credit(ConsumerKey, Credit, DeliveryCount, Drain),
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{send_command(ServerId, undefined, Cmd, normal, State), []}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
%% @doc Cancels a checkout with the rabbit_fifo queue for the consumer tag
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%%
2019-02-13 01:28:22 +08:00
%% This is a synchronous call. I.e. the call will block until the command
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% has been accepted by the ra process or it times out.
%%
%% @param ConsumerTag a unique tag to identify this particular consumer.
%% @param State The {@module} state.
%%
%% @returns `{ok, State}' or `{error | timeout, term()}'
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
-spec cancel_checkout(rabbit_types:ctag(), rabbit_queue_type:cancel_reason(), state()) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
{ok, state()} | {error | timeout, term()}.
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
cancel_checkout(ConsumerTag, Reason,
#state{consumers = Consumers} = State0)
when is_atom(Reason) ->
case Consumers of
#{ConsumerTag := #consumer{key = Cid}} ->
Servers = sorted_servers(State0),
ConsumerId = {ConsumerTag, self()},
State1 = send_pending(Cid, State0),
Cmd = rabbit_fifo:make_checkout(ConsumerId, Reason, #{}),
State = State1#state{consumers = maps:remove(ConsumerTag, Consumers)},
case try_process_command(Servers, Cmd, State) of
{ok, _, Leader} ->
{ok, State#state{leader = Leader}};
Err ->
Err
end;
_ ->
%% TODO: when we implement the `delete' checkout spec we could
%% fallback to that to make sure there is little chance a consumer
%% sticks around in the machine
{ok, State0}
end.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Purges all the messages from a rabbit_fifo queue and returns the number
%% of messages purged.
-spec purge(ra:server_id()) -> {ok, non_neg_integer()} | {error | timeout, term()}.
purge(Server) ->
case ra:process_command(Server, rabbit_fifo:make_purge(), ?COMMAND_TIMEOUT) of
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
{ok, {purge, Reply}, _} ->
{ok, Reply};
Err ->
Err
end.
-spec pending_size(state()) -> non_neg_integer().
pending_size(#state{pending = Pend}) ->
maps:size(Pend).
-spec num_cached_segments(state()) -> non_neg_integer().
num_cached_segments(#state{cached_segments = CachedSegments}) ->
case CachedSegments of
undefined ->
0;
{_, _, Cached} ->
ra_flru:size(Cached)
end.
-spec stat(ra:server_id()) ->
{ok, non_neg_integer(), non_neg_integer()}
| {error | timeout, term()}.
stat(Leader) ->
%% short timeout as we don't want to spend too long if it is going to
%% fail anyway
stat(Leader, 250).
-spec stat(ra:server_id(), non_neg_integer()) ->
{ok, non_neg_integer(), non_neg_integer()}
| {error | timeout, term()}.
stat(Leader, Timeout) ->
%% short timeout as we don't want to spend too long if it is going to
%% fail anyway
case ra:local_query(Leader, fun rabbit_fifo:query_stat/1, Timeout) of
{ok, {_, {R, C}}, _} -> {ok, R, C};
{error, _} = Error -> Error;
{timeout, _} = Error -> Error
end.
update_machine_state(Server, Conf) ->
case ra:process_command(Server, rabbit_fifo:make_update_config(Conf), ?COMMAND_TIMEOUT) of
{ok, ok, _} ->
ok;
Err ->
Err
end.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Handles incoming `ra_events'. Events carry both internal "bookeeping"
%% events emitted by the `ra' leader as well as `rabbit_fifo' emitted events such
%% as message deliveries. All ra events need to be handled by {@module}
%% to ensure bookeeping, resends and flow control is correctly handled.
%%
%% If the `ra_event' contains a `rabbit_fifo' generated message it will be returned
%% for further processing.
%%
%% Example:
%%
%% ```
%% receive
%% {ra_event, From, Evt} ->
%% case rabbit_fifo_client:handle_ra_event(From, Evt, State0) of
%% {internal, _Seq, State} -> State;
%% {{delivery, _ConsumerTag, Msgs}, State} ->
%% handle_messages(Msgs),
%% ...
%% end
%% end
%% '''
%%
%% @param QName Name of the queue.
%% @param From the {@link ra:server_id().} of the sending process.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @param Event the body of the `ra_event'.
%% @param State the current {@module} state.
%%
%% @returns
%% `{internal, AppliedCorrelations, State}' if the event contained an internally
%% handled event such as a notification and a correlation was included with
%% the command (e.g. in a call to `enqueue/3' the correlation terms are returned
%% here).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%%
%% `{RaFifoEvent, State}' if the event contained a client message generated by
%% the `rabbit_fifo' state machine such as a delivery.
%%
%% The type of `rabbit_fifo' client messages that can be received are:
%%
%% `{delivery, ConsumerTag, [{MsgId, {MsgHeader, Msg}}]}'
%%
%% <li>`ConsumerTag' the binary tag passed to {@link checkout/3.}</li>
%% <li>`MsgId' is a consumer scoped monotonically incrementing id that can be
%% used to {@link settle/3.} (roughly: AMQP 0.9.1 ack) message once finished
%% with them.</li>
-spec handle_ra_event(rabbit_amqqueue:name(), ra:server_id(),
ra_server_proc:ra_event_body(), state()) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{internal, Correlators :: [term()], rabbit_queue_type:actions(), state()} |
{rabbit_fifo:client_msg(), state()} | {eol, rabbit_queue_type:actions()}.
handle_ra_event(QName, Leader, {applied, Seqs},
#state{leader = OldLeader,
cfg = #cfg{soft_limit = SftLmt}} = State0) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{Corrs, ActionsRev, State1} = lists:foldl(fun seq_applied/2,
{[], [], State0#state{leader = Leader}},
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
Seqs),
%% if the leader has changed we need to resend any pending commands remaining
%% after the applied processing
State2 = if OldLeader =/= Leader ->
%% double check before resending as applied notifications
%% can arrive from old leaders in any order
case ra:members(Leader) of
{ok, _, ActualLeader}
when ActualLeader =/= OldLeader ->
%% there is a new leader
?LOG_DEBUG("~ts: Detected QQ leader change (applied) "
2025-07-16 06:32:20 +08:00
"from ~w to ~w, "
"resending ~b pending commands",
[?MODULE, OldLeader, ActualLeader,
maps:size(State1#state.pending)]),
resend_all_pending(State1#state{leader = ActualLeader});
_ ->
State1
end;
true ->
State1
end,
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
Actions0 = lists:reverse(ActionsRev),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
Actions = case Corrs of
[] ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
Actions0;
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
_ ->
%%TODO: consider using lists:foldr/3 above because
%% Corrs is returned in the wrong order here.
%% The wrong order does not matter much because the channel sorts the
%% sequence numbers before confirming to the client. But rabbit_fifo_client
%% is sequence numer agnostic: it handles any correlation terms.
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
[{settled, QName, Corrs} | Actions0]
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
end,
case map_size(State2#state.pending) < SftLmt of
true when State2#state.slow == true ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
% we have exited soft limit state
% send any unsent commands and cancel the time as
% TODO: really the timer should only be cancelled when the channel
% exits flow state (which depends on the state of all queues the
% channel is interacting with)
% but the fact the queue has just applied suggests
% it's ok to cancel here anyway
State3 = cancel_timer(State2#state{slow = false,
unsent_commands = #{}}),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
% build up a list of commands to issue
Commands = maps:fold(
fun (Cid, {Settled, Returns, Discards}, Acc) ->
add_command(Cid, settle, Settled,
add_command(Cid, return, Returns,
add_command(Cid, discard,
Discards, Acc)))
end, [], State2#state.unsent_commands),
ServerId = pick_server(State3),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% send all the settlements and returns
State = lists:foldl(fun (C, S0) ->
send_command(ServerId, undefined, C,
normal, S0)
end, State3, Commands),
{ok, State, [{unblock, cluster_name(State)} | Actions]};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
_ ->
{ok, State2, Actions}
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
end;
handle_ra_event(QName, From, {machine, Del}, State0)
when element(1, Del) == delivery ->
handle_delivery(QName, From, Del, State0);
Introduce outbound RabbitMQ internal AMQP flow control ## What? Introduce RabbitMQ internal flow control for messages sent to AMQP clients. Prior this PR, when an AMQP client granted a large amount of link credit (e.g. 100k) to the sending queue, the sending queue sent that amount of messages to the session process no matter what. This becomes problematic for memory usage when the session process cannot send out messages fast enough to the AMQP client, especially if 1. The writer proc cannot send fast enough. This can happen when the AMQP client does not receive fast enough and causes TCP back-pressure to the server. Or 2. The server session proc is limited by remote-incoming-window. Both scenarios are now added as test cases. Tests * tcp_back_pressure_rabbitmq_internal_flow_quorum_queue * tcp_back_pressure_rabbitmq_internal_flow_classic_queue cover scenario 1. Tests * incoming_window_closed_rabbitmq_internal_flow_quorum_queue * incoming_window_closed_rabbitmq_internal_flow_classic_queue cover scenario 2. This PR sends messages from queues to AMQP clients in a more controlled manner. To illustrate: ``` make run-broker PLUGINS="rabbitmq_management" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 4" observer_cli:start() mq ``` where `mq` sorts by message queue length. Create a stream: ``` deps/rabbitmq_management/bin/rabbitmqadmin declare queue name=s1 queue_type=stream durable=true ``` Next, send and receive from the Stream via AMQP. Grant a large number of link credit to the sending stream: ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT bash-5.1# quiver //host.docker.internal//queue/s1 --durable -d 30s --credit 100000 ``` **Before** to this PR: ``` RESULTS Count ............................................... 100,696 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 120,422 messages/s Receiver rate ......................................... 3,363 messages/s End-to-end rate ....................................... 3,359 messages/s ``` We observe that all 100k link credit worth of messages are buffered in the writer proc's mailbox: ``` |No | Pid | MsgQueue |Name or Initial Call | Memory | Reductions |Current Function | |1 |<0.845.0> |100001 |rabbit_amqp_writer:init/1 | 126.0734 MB| 466633491 |prim_inet:send/5 | ``` **After** to this PR: ``` RESULTS Count ............................................. 2,973,440 messages Duration ............................................... 30.0 seconds Sender rate ......................................... 123,322 messages/s Receiver rate ........................................ 99,250 messages/s End-to-end rate ...................................... 99,148 messages/s ``` We observe that the message queue lengths of both writer and session procs are low. ## How? Our goal is to have queues send out messages in a controlled manner without overloading RabbitMQ itself. We want RabbitMQ internal flow control between: ``` AMQP writer proc <--- session proc <--- queue proc ``` A similar concept exists for classic queues sending via AMQP 0.9.1. We want an approach that applies to AMQP and works generic for all queue types. For the interaction between AMQP writer proc and session proc we use a simple credit based approach reusing module `credit_flow`. For the interaction between session proc and queue proc, the following options exist: ### Option 1 The session process provides expliclity feedback to the queue after it has sent N messages. This approach is implemented in https://github.com/ansd/rabbitmq-server/tree/amqp-flow-control-poc-1 and works well. A new `rabbit_queue_type:sent/4` API was added which lets the queue proc know that it can send further messages to the session proc. Pros: * Will work equally well for AMQP 0.9.1, e.g. when quorum queues send messages in auto ack mode to AMQP 0.9.1 clients. * Simple for the session proc Cons: * Sligthly added complexity in every queue type implementation * Multiple Ra commands (settle, credit, sent) to decide when a quorum queue sends more messages. ### Option 2 A dual link approach where two AMQP links exists between ``` AMQP client <---link--> session proc <---link---> queue proc ``` When the client grants a large amount of credits, the session proc will top up credits to the queue proc periodically in smaller batches. Pros: * No queue type modifications required. * Re-uses AMQP link flow control Cons: * Significant added complexity in the session proc. A client can dynamically decrease or increase credits and dynamically change the drain mode while the session tops up credit to the queue. ### Option 3 Credit is a 32 bit unsigned integer. The spec mandates that the receiver independently chooses a credit. Nothing in the spec prevents the receiver to choose a credit of 1 billion. However the credit value is merely a **maximum**: > The link-credit variable defines the current maximum legal amount that the delivery-count can be increased by. Therefore, the server is not required to send all available messages to this receiver. For delivery-count: > Only the sender MAY independently modify this field. "independently" could be interpreted as the sender could add to the delivery-count irrespective of what the client chose for drain and link-credit. Option 3: The queue proc could at credit time already consume credit and advance the delivery-count if credit is too large before checking out any messages. For example if credit is 100k, but the queue only wants to send 1k, the queue could consume 99k of credits and advance the delivery-count, and subsequently send maximum 1k messages. If the queue advanced the delivery-count, RabbitMQ must send a FLOW to the receiver, otherwise the receiver wouldn’t know that it ran out of link-credit. Pros: * Very simple Cons: * Possibly unexpected behaviour for receiving AMQP clients * Possibly poor end-to-end throughput in auto-ack mode because the queue would send a batch of messages followed by a FLOW containing the advanced delivery-count. Only therafter the client will learn that it ran out of credits and top-up again. This feels like synchronously pulling a batch of messages. In contrast, option 2 sends out more messages as soon as the previous messages left RabbitMQ without requiring again a credit top up from the receiver. * drain mode with large credits requires the queue to send all available messages and only thereafter advance the delivery-count. Therefore, drain mode breaks option 3 somewhat. ### Option 4 Session proc drops message payload when its outgoing-pending queue gets too large and re-reads payloads from the queue once the message can be sent (see `get_checked_out` Ra command for quorum queues). Cons: * Would need to be implemented for every queue type, especially classic queues * Doesn't limit the amount of message metadata in the session proc's outgoing-pending queue ### Decision: Option 2 This commit implements option 2 to avoid any queue type modification. At most one credit request is in-flight between session process and queue process for a given queue consumer. If the AMQP client sends another FLOW in between, the session proc stashes the FLOW until it processes the previous credit reply. A delivery is only sent from the outgoing-pending queue if the session proc is not blocked by 1. writer proc, or 2. remote-incoming-window The credit reply is placed into the outgoing-pending queue. This ensures that the session proc will only top up the next batch of credits if sufficient messages were sent out to the writer proc. A future commit could additionally have each queue limit the number of unacked messages for a given AMQP consumer, or alternatively make use of session outgoing-window.
2024-05-18 17:11:12 +08:00
handle_ra_event(_QName, _From, {machine, Action}, State)
when element(1, Action) =:= credit_reply orelse
element(1, Action) =:= credit_reply_v1 ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
{ok, State, [Action]};
handle_ra_event(_QName, _, {machine, {queue_status, Status}},
#state{} = State) ->
%% just set the queue status
2020-10-02 20:37:25 +08:00
{ok, State#state{queue_status = Status}, []};
handle_ra_event(QName, Leader, {machine, leader_change},
#state{leader = OldLeader,
pending = Pending} = State0) ->
%% we need to update leader
%% and resend any pending commands
?LOG_DEBUG("~ts: ~s Detected QQ leader change from ~w to ~w, "
2025-07-16 06:32:20 +08:00
"resending ~b pending commands",
[rabbit_misc:rs(QName), ?MODULE, OldLeader,
Leader, maps:size(Pending)]),
State = resend_all_pending(State0#state{leader = Leader}),
{ok, State, []};
handle_ra_event(_QName, _From, {rejected, {not_leader, Leader, _Seq}},
#state{leader = Leader} = State) ->
{ok, State, []};
handle_ra_event(QName, _From, {rejected, {not_leader, Leader, _Seq}},
#state{leader = OldLeader,
pending = Pending} = State0) ->
?LOG_DEBUG("~ts: ~s Detected QQ leader change (rejection) from ~w to ~w, "
2025-07-16 06:32:20 +08:00
"resending ~b pending commands",
[rabbit_misc:rs(QName), ?MODULE, OldLeader,
Leader, maps:size(Pending)]),
State = resend_all_pending(State0#state{leader = Leader}),
{ok, cancel_timer(State), []};
handle_ra_event(_QName, _From,
{rejected, {not_leader, _UndefinedMaybe, _Seq}}, State0) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
% TODO: how should these be handled? re-sent on timer or try random
{ok, State0, []};
handle_ra_event(QName, _, timeout, #state{cfg = #cfg{servers = Servers}} = State0) ->
case find_leader(Servers) of
undefined ->
%% still no leader, set the timer again
{ok, set_timer(QName, State0), []};
Leader ->
State = resend_all_pending(State0#state{leader = Leader}),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{ok, State, []}
end;
handle_ra_event(QName, Leader, close_cached_segments,
#state{cached_segments = CachedSegments} = State) ->
{ok,
case CachedSegments of
undefined ->
%% timer didn't get cancelled so just ignore this
State;
{_TRef, Last, Cache} ->
case now_ms() > Last + ?CACHE_SEG_TIMEOUT of
true ->
?LOG_DEBUG("~ts: closing_cached_segments",
[rabbit_misc:rs(QName)]),
%% its been long enough, evict all
_ = ra_flru:evict_all(Cache),
State#state{cached_segments = undefined};
false ->
%% set another timer
Ref = erlang:send_after(?CACHE_SEG_TIMEOUT, self(),
{'$gen_cast',
{queue_event, QName,
{Leader, close_cached_segments}}}),
State#state{cached_segments = {Ref, Last, Cache}}
end
end, []};
handle_ra_event(_QName, _Leader, {machine, eol}, State) ->
{eol, [{unblock, cluster_name(State)}]}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
-spec close(rabbit_fifo_client:state()) -> ok.
close(#state{cached_segments = undefined}) ->
ok;
close(#state{cached_segments = {_, _, Flru}}) ->
_ = ra_flru:evict_all(Flru),
ok.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% @doc Attempts to enqueue a message using cast semantics. This provides no
%% guarantees or retries if the message fails to achieve consensus or if the
%% servers sent to happens not to be available. If the message is sent to a
%% follower it will attempt the deliver it to the leader, if known. Else it will
%% drop the messages.
%%
%% NB: only use this for non-critical enqueues where a full rabbit_fifo_client state
%% cannot be maintained.
%%
%% @param CusterId the cluster id.
%% @param Servers the known servers in the cluster.
%% @param Msg the message to enqueue.
%%
%% @returns `ok'
-spec untracked_enqueue([ra:server_id()], term()) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
ok.
untracked_enqueue([ServerId | _], Msg) ->
Cmd = rabbit_fifo:make_enqueue(undefined, undefined, Msg),
ok = ra:pipeline_command(ServerId, Cmd),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
ok.
%% Internal
try_process_command([Server | Rem], Cmd,
#state{cfg = #cfg{timeout = Timeout}} = State) ->
case ra:process_command(Server, Cmd, Timeout) of
{ok, _, _} = Res ->
Res;
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Err when length(Rem) =:= 0 ->
Err;
_ ->
try_process_command(Rem, Cmd, State)
end.
seq_applied({Seq, Response},
{Corrs, Actions0, #state{} = State0}) ->
%% sequences aren't guaranteed to be applied in order as enqueues are
%% low priority commands and may be overtaken by others with a normal priority.
{Actions, State} = maybe_add_action(Response, Actions0, State0),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
case maps:take(Seq, State#state.pending) of
{{undefined, _}, Pending} ->
{Corrs, Actions, State#state{pending = Pending}};
{{Corr, _}, Pending}
when Response /= not_enqueued ->
{[Corr | Corrs], Actions, State#state{pending = Pending}};
_ ->
{Corrs, Actions, State}
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
end;
seq_applied(_Seq, Acc) ->
Acc.
maybe_add_action(ok, Acc, State) ->
{Acc, State};
maybe_add_action(not_enqueued, Acc, State) ->
{Acc, State};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
maybe_add_action({multi, Actions}, Acc0, State0) ->
lists:foldl(fun (Act, {Acc, State}) ->
maybe_add_action(Act, Acc, State)
end, {Acc0, State0}, Actions);
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
maybe_add_action({send_drained, {Tag, Credit}}, Acc, State0) ->
%% This function clause should be deleted when
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
%% feature flag rabbitmq_4.0.0 becomes required.
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
State = add_delivery_count(Credit, Tag, State0),
Action = {credit_reply_v1, Tag, Credit, _Avail = 0, _Drain = true},
{[Action | Acc], State};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
maybe_add_action(Action, Acc, State) ->
%% anything else is assumed to be an action
{[Action | Acc], State}.
% resends a command with a new sequence number
resend(OldSeq, #state{pending = Pending0, leader = Leader} = State) ->
case maps:take(OldSeq, Pending0) of
{{Corr, Cmd}, Pending} ->
%% resends aren't subject to flow control here
resend_command(Leader, Corr, Cmd, State#state{pending = Pending});
error ->
State
end.
2020-07-10 21:25:17 +08:00
resend_all_pending(#state{pending = Pend} = State) ->
Seqs = lists:sort(maps:keys(Pend)),
lists:foldl(fun resend/2, State, Seqs).
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
maybe_auto_ack(true, Deliver, State0) ->
%% manual ack is enabled
{ok, State0, [Deliver]};
maybe_auto_ack(false, {deliver, Tag, _Ack, Msgs} = Deliver, State0) ->
%% we have to auto ack these deliveries
MsgIds = [I || {_, _, I, _, _} <- Msgs],
{State, Actions} = settle(Tag, MsgIds, State0),
{ok, State, [Deliver] ++ Actions}.
handle_delivery(QName, Leader, {delivery, Tag, [{FstId, _} | _] = IdMsgs},
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#state{consumers = CDels0} = State0)
when is_map_key(Tag, CDels0) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
QRef = qref(Leader),
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
{LastId, _} = lists:last(IdMsgs),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
Consumer = #consumer{ack = Ack} = maps:get(Tag, CDels0),
%% format as a deliver action
Del = {deliver, Tag, Ack, transform_msgs(QName, QRef, IdMsgs)},
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
%% TODO: remove potential default allocation
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
case Consumer of
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
#consumer{last_msg_id = Prev} = C
when Prev =:= undefined orelse FstId =:= Prev+1 ->
%% Prev =:= undefined is a special case where a checkout was done
%% for a previously cancelled consumer that still had pending messages
%% In this case we can't reliably know what the next expected message
%% id should be so have to accept whatever message comes next
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
maybe_auto_ack(Ack, Del,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
State0#state{consumers =
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
update_consumer(Tag, LastId,
length(IdMsgs), C,
CDels0)});
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
#consumer{last_msg_id = Prev} = C
when FstId > Prev+1 ->
NumMissing = FstId - Prev + 1,
%% there may actually be fewer missing messages returned than expected
%% This can happen when a node the channel is on gets disconnected
%% from the node the leader is on and then reconnected afterwards.
%% When the node is disconnected the leader will return all checked
%% out messages to the main queue to ensure they don't get stuck in
%% case the node never comes back.
case get_missing_deliveries(State0, Prev+1, FstId-1, Tag) of
{protocol_error, _, _, _} = Err ->
Err;
Missing ->
XDel = {deliver, Tag, Ack, transform_msgs(QName, QRef,
Missing ++ IdMsgs)},
maybe_auto_ack(Ack, XDel,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
State0#state{consumers =
update_consumer(Tag, LastId,
length(IdMsgs) + NumMissing,
C, CDels0)})
end;
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
#consumer{last_msg_id = Prev}
when FstId =< Prev ->
case lists:dropwhile(fun({Id, _}) -> Id =< Prev end, IdMsgs) of
[] ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{ok, State0, []};
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
IdMsgs2 ->
handle_delivery(QName, Leader, {delivery, Tag, IdMsgs2}, State0)
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
end;
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
C when FstId =:= 0 ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
% the very first delivery
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
maybe_auto_ack(Ack, Del,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
State0#state{consumers =
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
update_consumer(Tag, LastId,
length(IdMsgs),
C#consumer{last_msg_id = LastId},
CDels0)})
end;
handle_delivery(_QName, _Leader, {delivery, Tag, [_ | _] = IdMsgs},
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
#state{consumers = CDels0} = State0)
when not is_map_key(Tag, CDels0) ->
%% Note:
%% https://github.com/rabbitmq/rabbitmq-server/issues/3729
%% If the consumer is no longer in the deliveries map,
%% we should return all messages.
MsgIntIds = [Id || {Id, _} <- IdMsgs],
{State1, Deliveries} = return(Tag, MsgIntIds, State0),
{ok, State1, Deliveries};
handle_delivery(QName, Leader, {delivery, Tag, ReadPlan, Msgs},
#state{cached_segments = CachedSegments} = State) ->
{TRef, Cached0} = case CachedSegments of
undefined ->
{undefined, undefined};
{R, _, C} ->
{R, C}
end,
{MsgIds, Cached1} = rabbit_fifo:exec_read(Cached0, ReadPlan, Msgs),
%% if there are cached segments after a read and there
%% is no current timer set, set a timer
%% send a message to evict cache after some time
Cached = case ra_flru:size(Cached1) > 0 of
true when TRef == undefined ->
Ref = erlang:send_after(?CACHE_SEG_TIMEOUT, self(),
{'$gen_cast',
{queue_event, QName,
{Leader, close_cached_segments}}}),
{Ref, now_ms(), Cached1};
true ->
{TRef, now_ms(), Cached1};
false when is_reference(TRef) ->
%% the time is (potentially) alive and may as well be
%% cancelled here
_ = erlang:cancel_timer(TRef, [{async, true},
{info, false}]),
undefined;
false ->
undefined
end,
handle_delivery(QName, Leader, {delivery, Tag, MsgIds},
State#state{cached_segments = Cached}).
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
transform_msgs(QName, QRef, Msgs) ->
lists:map(
fun({MsgId, {MsgHeader, Msg0}}) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
{Msg, Redelivered} = add_delivery_count_header(Msg0, MsgHeader),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{QName, QRef, MsgId, Redelivered, Msg}
end, Msgs).
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
update_consumer(Tag, LastId, DelCntIncr, Consumer, Consumers) ->
D = case Consumer#consumer.delivery_count of
credit_api_v2 -> credit_api_v2;
{credit_api_v1, Count} -> {credit_api_v1, Count + DelCntIncr}
end,
maps:update(Tag,
Consumer#consumer{last_msg_id = LastId,
delivery_count = D},
Consumers).
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
add_delivery_count(DelCntIncr, Tag, #state{consumers = CDels0} = State) ->
Support AMQP 1.0 natively ## What Similar to Native MQTT in #5895, this commits implements Native AMQP 1.0. By "native", we mean do not proxy via AMQP 0.9.1 anymore. ## Why Native AMQP 1.0 comes with the following major benefits: 1. Similar to Native MQTT, this commit provides better throughput, latency, scalability, and resource usage for AMQP 1.0. See https://blog.rabbitmq.com/posts/2023/03/native-mqtt for native MQTT improvements. See further below for some benchmarks. 2. Since AMQP 1.0 is not limited anymore by the AMQP 0.9.1 protocol, this commit allows implementing more AMQP 1.0 features in the future. Some features are already implemented in this commit (see next section). 3. Simpler, better understandable, and more maintainable code. Native AMQP 1.0 as implemented in this commit has the following major benefits compared to AMQP 0.9.1: 4. Memory and disk alarms will only stop accepting incoming TRANSFER frames. New connections can still be created to consume from RabbitMQ to empty queues. 5. Due to 4. no need anymore for separate connections for publishers and consumers as we currently recommended for AMQP 0.9.1. which potentially halves the number of physical TCP connections. 6. When a single connection sends to multiple target queues, a single slow target queue won't block the entire connection. Publisher can still send data quickly to all other target queues. 7. A publisher can request whether it wants publisher confirmation on a per-message basis. In AMQP 0.9.1 publisher confirms are configured per channel only. 8. Consumers can change their "prefetch count" dynamically which isn't possible in our AMQP 0.9.1 implementation. See #10174 9. AMQP 1.0 is an extensible protocol This commit also fixes dozens of bugs present in the AMQP 1.0 plugin in RabbitMQ 3.x - most of which cannot be backported due to the complexity and limitations of the old 3.x implementation. This commit contains breaking changes and is therefore targeted for RabbitMQ 4.0. ## Implementation details 1. Breaking change: With Native AMQP, the behaviour of ``` Convert AMQP 0.9.1 message headers to application properties for an AMQP 1.0 consumer amqp1_0.convert_amqp091_headers_to_app_props = false | true (default false) Convert AMQP 1.0 Application Properties to AMQP 0.9.1 headers amqp1_0.convert_app_props_to_amqp091_headers = false | true (default false) ``` will break because we always convert according to the message container conversions. For example, AMQP 0.9.1 x-headers will go into message-annotations instead of application properties. Also, `false` won’t be respected since we always convert the headers with message containers. 2. Remove rabbit_queue_collector rabbit_queue_collector is responsible for synchronously deleting exclusive queues. Since the AMQP 1.0 plugin never creates exclusive queues, rabbit_queue_collector doesn't need to be started in the first place. This will save 1 Erlang process per AMQP 1.0 connection. 3. 7 processes per connection + 1 process per session in this commit instead of 7 processes per connection + 15 processes per session in 3.x Supervision hierarchy got re-designed. 4. Use 1 writer process per AMQP 1.0 connection AMQP 0.9.1 uses a separate rabbit_writer Erlang process per AMQP 0.9.1 channel. Prior to this commit, AMQP 1.0 used a separate rabbit_amqp1_0_writer process per AMQP 1.0 session. Advantage of single writer proc per session (prior to this commit): * High parallelism for serialising packets if multiple sessions within a connection write heavily at the same time. This commit uses a single writer process per AMQP 1.0 connection that is shared across all AMQP 1.0 sessions. Advantages of single writer proc per connection (this commit): * Lower memory usage with hundreds of thousands of AMQP 1.0 sessions * Less TCP and IP header overhead given that the single writer process can accumulate across all sessions bytes before flushing the socket. In other words, this commit decides that a reader / writer process pair per AMQP 1.0 connection is good enough for bi-directional TRANSFER flows. Having a writer per session is too heavy. We still ensure high throughput by having separate reader, writer, and session processes. 5. Transform rabbit_amqp1_0_writer into gen_server Why: Prior to this commit, when clicking on the AMQP 1.0 writer process in observer, the process crashed. Instead of handling all these debug messages of the sys module, it's better to implement a gen_server. There is no advantage of using a special OTP process over gen_server for the AMQP 1.0 writer. gen_server also provides cleaner format status output. How: Message callbacks return a timeout of 0. After all messages in the inbox are processed, the timeout message is handled by flushing any pending bytes. 6. Remove stats timer from writer AMQP 1.0 connections haven't emitted any stats previously. 7. When there are contiguous queue confirmations in the session process mailbox, batch them. When the confirmations are sent to the publisher, a single DISPOSITION frame is sent for contiguously confirmed delivery IDs. This approach should be good enough. However it's sub optimal in scenarios where contiguous delivery IDs that need confirmations are rare, for example: * There are multiple links in the session with different sender settlement modes and sender publishes across these links interleaved. * sender settlement mode is mixed and sender publishes interleaved settled and unsettled TRANSFERs. 8. Introduce credit API v2 Why: The AMQP 0.9.1 credit extension which is to be removed in 4.0 was poorly designed since basic.credit is a synchronous call into the queue process blocking the entire AMQP 1.0 session process. How: Change the interactions between queue clients and queue server implementations: * Clients only request a credit reply if the FLOW's `echo` field is set * Include all link flow control state held by the queue process into a new credit_reply queue event: * `available` after the queue sends any deliveries * `link-credit` after the queue sends any deliveries * `drain` which allows us to combine the old queue events send_credit_reply and send_drained into a single new queue event credit_reply. * Include the consumer tag into the credit_reply queue event such that the AMQP 1.0 session process can process any credit replies asynchronously. Link flow control state `delivery-count` also moves to the queue processes. The new interactions are hidden behind feature flag credit_api_v2 to allow for rolling upgrades from 3.13 to 4.0. 9. Use serial number arithmetic in quorum queues and session process. 10. Completely bypass the rabbit_limiter module for AMQP 1.0 flow control. The goal is to eventually remove the rabbit_limiter module in 4.0 since AMQP 0.9.1 global QoS will be unsupported in 4.0. This commit lifts the AMQP 1.0 link flow control logic out of rabbit_limiter into rabbit_queue_consumers. 11. Fix credit bug for streams: AMQP 1.0 settlements shouldn't top up link credit, only FLOW frames should top up link credit. 12. Allow sender settle mode unsettled for streams since AMQP 1.0 acknowledgements to streams are no-ops (currently). 13. Fix AMQP 1.0 client bugs Auto renewing credits should not be related to settling TRANSFERs. Remove field link_credit_unsettled as it was wrong and confusing. Prior to this commit auto renewal did not work when the sender uses sender settlement mode settled. 14. Fix AMQP 1.0 client bugs The wrong outdated Link was passed to function auto_flow/2 15. Use osiris chunk iterator Only hold messages of uncompressed sub batches in memory if consumer doesn't have sufficient credits. Compressed sub batches are skipped for non Stream protocol consumers. 16. Fix incoming link flow control Always use confirms between AMQP 1.0 queue clients and queue servers. As already done internally by rabbit_fifo_client and rabbit_stream_queue, use confirms for classic queues as well. 17. Include link handle into correlation when publishing messages to target queues such that session process can correlate confirms from target queues to incoming links. 18. Only grant more credits to publishers if publisher hasn't sufficient credits anymore and there are not too many unconfirmed messages on the link. 19. Completely ignore `block` and `unblock` queue actions and RabbitMQ credit flow between classic queue process and session process. 20. Link flow control is independent between links. A client can refer to a queue or to an exchange with multiple dynamically added target queues. Multiple incoming links can also fan in to the same queue. However the link topology looks like, this commit ensures that each link is only granted more credits if that link isn't overloaded. 21. A connection or a session can send to many different queues. In AMQP 0.9.1, a single slow queue will lead to the entire channel, and then entire connection being blocked. This commit makes sure that a single slow queue from one link won't slow down sending on other links. For example, having link A sending to a local classic queue and link B sending to 5 replica quorum queue, link B will naturally grant credits slower than link A. So, despite the quorum queue being slower in confirming messages, the same AMQP 1.0 connection and session can still pump data very fast into the classic queue. 22. If cluster wide memory or disk alarm occurs. Each session sends a FLOW with incoming-window to 0 to sending client. If sending clients don’t obey, force disconnect the client. If cluster wide memory alarm clears: Each session resumes with a FLOW defaulting to initial incoming-window. 23. All operations apart of publishing TRANSFERS to RabbitMQ can continue during cluster wide alarms, specifically, attaching consumers and consuming, i.e. emptying queues. There is no need for separate AMQP 1.0 connections for publishers and consumers as recommended in our AMQP 0.9.1 implementation. 24. Flow control summary: * If queue becomes bottleneck, that’s solved by slowing down individual sending links (AMQP 1.0 link flow control). * If session becomes bottleneck (more unlikely), that’s solved by AMQP 1.0 session flow control. * If connection becomes bottleneck, it naturally won’t read fast enough from the socket causing TCP backpressure being applied. Nowhere will RabbitMQ internal credit based flow control (i.e. module credit_flow) be used on the incoming AMQP 1.0 message path. 25. Register AMQP sessions Prefer local-only pg over our custom pg_local implementation as pg is a better process group implementation than pg_local. pg_local was identified as bottleneck in tests where many MQTT clients were disconnected at once. 26. Start a local-only pg when Rabbit boots: > A scope can be kept local-only by using a scope name that is unique cluster-wide, e.g. the node name: > pg:start_link(node()). Register AMQP 1.0 connections and sessions with pg. In future we should remove pg_local and instead use the new local-only pg for all registered processes such as AMQP 0.9.1 connections and channels. 27. Requeue messages if link detached Although the spec allows to settle delivery IDs on detached links, RabbitMQ does not respect the 'closed' field of the DETACH frame and therefore handles every DETACH frame as closed. Since the link is closed, we expect every outstanding delivery to be requeued. In addition to consumer cancellation, detaching a link therefore causes in flight deliveries to be requeued. Note that this behaviour is different from merely consumer cancellation in AMQP 0.9.1: "After a consumer is cancelled there will be no future deliveries dispatched to it. Note that there can still be "in flight" deliveries dispatched previously. Cancelling a consumer will neither discard nor requeue them." [https://www.rabbitmq.com/consumers.html#unsubscribing] An AMQP receiver can first drain, and then detach to prevent "in flight" deliveries 28. Init AMQP session with BEGIN frame Similar to how there can't be an MQTT processor without a CONNECT frame, there can't be an AMQP session without a BEGIN frame. This allows having strict dialyzer types for session flow control fields (i.e. not allowing 'undefined'). 29. Move serial_number to AMQP 1.0 common lib such that it can be used by both AMQP 1.0 server and client 30. Fix AMQP client to do serial number arithmetic. 31. AMQP client: Differentiate between delivery-id and transfer-id for better understandability. 32. Fix link flow control in classic queues This commit fixes ``` java -jar target/perf-test.jar -ad false -f persistent -u cq -c 3000 -C 1000000 -y 0 ``` followed by ``` ./omq -x 0 amqp -T /queue/cq -D 1000000 --amqp-consumer-credits 2 ``` Prior to this commit, (and on RabbitMQ 3.x) the consuming would halt after around 8 - 10,000 messages. The bug was that in flight messages from classic queue process to session process were not taken into account when topping up credit to the classic queue process. Fixes #2597 The solution to this bug (and a much cleaner design anyway independent of this bug) is that queues should hold all link flow control state including the delivery-count. Hence, when credit API v2 is used the delivery-count will be held by the classic queue process, quorum queue process, and stream queue client instead of managing the delivery-count in the session. 33. The double level crediting between (a) session process and rabbit_fifo_client, and (b) rabbit_fifo_client and rabbit_fifo was removed. Therefore, instead of managing 3 separate delivery-counts (i. session, ii. rabbit_fifo_client, iii. rabbit_fifo), only 1 delivery-count is used in rabbit_fifo. This is a big simplification. 34. This commit fixes quorum queues without bumping the machine version nor introducing new rabbit_fifo commands. Whether credit API v2 is used is solely determined at link attachment time depending on whether feature flag credit_api_v2 is enabled. Even when that feature flag will be enabled later on, this link will keep using credit API v1 until detached (or the node is shut down). Eventually, after feature flag credit_api_v2 has been enabled and a subsequent rolling upgrade, all links will use credit API v2. This approach is safe and simple. The 2 alternatives to move delivery-count from the session process to the queue processes would have been: i. Explicit feature flag credit_api_v2 migration function * Can use a gen_server:call and only finish migration once all delivery-counts were migrated. Cons: * Extra new message format just for migration is required. * Risky as migration will fail if a target queue doesn’t reply. ii. Session always includes DeliveryCountSnd when crediting to the queue: Cons: * 2 delivery counts will be hold simultaneously in session proc and queue proc; could be solved by deleting the session proc’s delivery-count for credit-reply * What happens if the receiver doesn’t provide credit for a very long time? Is that a problem? 35. Support stream filtering in AMQP 1.0 (by @acogoluegnes) Use the x-stream-filter-value message annotation to carry the filter value in a published message. Use the rabbitmq:stream-filter and rabbitmq:stream-match-unfiltered filters when creating a receiver that wants to filter out messages from a stream. 36. Remove credit extension from AMQP 0.9.1 client 37. Support maintenance mode closing AMQP 1.0 connections. 38. Remove AMQP 0.9.1 client dependency from AMQP 1.0 implementation. 39. Move AMQP 1.0 plugin to the core. AMQP 1.0 is enabled by default. The old rabbitmq_amqp1_0 plugin will be kept as a no-op plugin to prevent deployment tools from failing that execute: ``` rabbitmq-plugins enable rabbitmq_amqp1_0 rabbitmq-plugins disable rabbitmq_amqp1_0 ``` 40. Breaking change: Remove CLI command `rabbitmqctl list_amqp10_connections`. Instead, list both AMQP 0.9.1 and AMQP 1.0 connections in `list_connections`: ``` rabbitmqctl list_connections protocol Listing connections ... protocol {1, 0} {0,9,1} ``` ## Benchmarks ### Throughput & Latency Setup: * Single node Ubuntu 22.04 * Erlang 26.1.1 Start RabbitMQ: ``` make run-broker PLUGINS="rabbitmq_management rabbitmq_amqp1_0" FULL=1 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+S 3" ``` Predeclare durable classic queue cq1, durable quorum queue qq1, durable stream queue sq1. Start client: https://github.com/ssorj/quiver https://hub.docker.com/r/ssorj/quiver/tags (digest 453a2aceda64) ``` docker run -it --rm --add-host host.docker.internal:host-gateway ssorj/quiver:latest bash-5.1# quiver --version quiver 0.4.0-SNAPSHOT ``` 1. Classic queue ``` quiver //host.docker.internal//amq/queue/cq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ............................................... 73.8 seconds Sender rate .......................................... 13,548 messages/s Receiver rate ........................................ 13,547 messages/s End-to-end rate ...................................... 13,547 messages/s Latencies by percentile: 0% ........ 0 ms 90.00% ........ 9 ms 25% ........ 2 ms 99.00% ....... 14 ms 50% ........ 4 ms 99.90% ....... 17 ms 100% ....... 26 ms 99.99% ....... 24 ms ``` RabbitMQ 3.x (main branch as of 30 January 2024): ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 6 73.6 2.1 3,217 1,607 0 8.0 511 4.1 163,580 16,367 2 74.1 4.1 3,217 0 0 8.0 0 6.1 229,114 32,767 3 74.1 6.1 3,217 0 0 8.0 0 8.1 261,880 16,367 2 74.1 8.1 67,874 32,296 8 8.2 7,662 10.1 294,646 16,367 2 74.1 10.1 67,874 0 0 8.2 0 12.1 360,180 32,734 3 74.1 12.1 67,874 0 0 8.2 0 14.1 392,946 16,367 3 74.1 14.1 68,604 365 0 8.2 12,147 16.1 458,480 32,734 3 74.1 16.1 68,604 0 0 8.2 0 18.1 491,246 16,367 2 74.1 18.1 68,604 0 0 8.2 0 20.1 556,780 32,767 4 74.1 20.1 68,604 0 0 8.2 0 22.1 589,546 16,375 2 74.1 22.1 68,604 0 0 8.2 0 receiver timed out 24.1 622,312 16,367 2 74.1 24.1 68,604 0 0 8.2 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/cq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-otujr23y' returned non-zero exit status 1. ``` 2. Quorum queue: ``` quiver //host.docker.internal//amq/queue/qq1 --durable --count 1m --duration 10m --body-size 12 --credit 1000 ``` This commit: ``` Count ............................................. 1,000,000 messages Duration .............................................. 101.4 seconds Sender rate ........................................... 9,867 messages/s Receiver rate ......................................... 9,868 messages/s End-to-end rate ....................................... 9,865 messages/s Latencies by percentile: 0% ....... 11 ms 90.00% ....... 23 ms 25% ....... 15 ms 99.00% ....... 28 ms 50% ....... 18 ms 99.90% ....... 33 ms 100% ....... 49 ms 99.99% ....... 47 ms ``` RabbitMQ 3.x: ``` ---------------------- Sender ----------------------- --------------------- Receiver ---------------------- -------- Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Time [s] Count [m] Rate [m/s] CPU [%] RSS [M] Lat [ms] ----------------------------------------------------- ----------------------------------------------------- -------- 2.1 130,814 65,342 9 69.9 2.1 18,430 9,206 5 7.6 1,221 4.1 163,580 16,375 5 70.2 4.1 18,867 218 0 7.6 2,168 6.1 229,114 32,767 6 70.2 6.1 18,867 0 0 7.6 0 8.1 294,648 32,734 7 70.2 8.1 18,867 0 0 7.6 0 10.1 360,182 32,734 6 70.2 10.1 18,867 0 0 7.6 0 12.1 425,716 32,767 6 70.2 12.1 18,867 0 0 7.6 0 receiver timed out 14.1 458,482 16,367 5 70.2 14.1 18,867 0 0 7.6 0 quiver: error: PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. Traceback (most recent call last): File "/usr/local/lib/quiver/python/quiver/pair.py", line 144, in run _plano.wait(receiver, check=True) File "/usr/local/lib/quiver/python/plano/main.py", line 1243, in wait raise PlanoProcessError(proc) plano.main.PlanoProcessError: Command 'quiver-arrow receive //host.docker.internal//amq/queue/qq1 --impl qpid-proton-c --duration 10m --count 1m --rate 0 --body-size 12 --credit 1000 --transaction-size 0 --timeout 10 --durable --output /tmp/quiver-b1gcup43' returned non-zero exit status 1. ``` 3. Stream: ``` quiver-arrow send //host.docker.internal//amq/queue/sq1 --durable --count 1m -d 10m --summary --verbose ``` This commit: ``` Count ............................................. 1,000,000 messages Duration ................................................ 8.7 seconds Message rate ........................................ 115,154 messages/s ``` RabbitMQ 3.x: ``` Count ............................................. 1,000,000 messages Duration ............................................... 21.2 seconds Message rate ......................................... 47,232 messages/s ``` ### Memory usage Start RabbitMQ: ``` ERL_MAX_PORTS=3000000 RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS="+P 3000000 +S 6" make run-broker PLUGINS="rabbitmq_amqp1_0" FULL=1 RABBITMQ_CONFIG_FILE="rabbitmq.conf" ``` ``` /bin/cat rabbitmq.conf tcp_listen_options.sndbuf = 2048 tcp_listen_options.recbuf = 2048 vm_memory_high_watermark.relative = 0.95 vm_memory_high_watermark_paging_ratio = 0.95 loopback_users = none ``` Create 50k connections with 2 sessions per connection, i.e. 100k session in total: ```go package main import ( "context" "log" "time" "github.com/Azure/go-amqp" ) func main() { for i := 0; i < 50000; i++ { conn, err := amqp.Dial(context.TODO(), "amqp://nuc", &amqp.ConnOptions{SASLType: amqp.SASLTypeAnonymous()}) if err != nil { log.Fatal("dialing AMQP server:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } _, err = conn.NewSession(context.TODO(), nil) if err != nil { log.Fatal("creating AMQP session:", err) } } log.Println("opened all connections") time.Sleep(5 * time.Hour) } ``` This commit: ``` erlang:memory(). [{total,4586376480}, {processes,4025898504}, {processes_used,4025871040}, {system,560477976}, {atom,1048841}, {atom_used,1042841}, {binary,233228608}, {code,21449982}, {ets,108560464}] erlang:system_info(process_count). 450289 ``` 7 procs per connection + 1 proc per session. (7 + 2*1) * 50,000 = 450,000 procs RabbitMQ 3.x: ``` erlang:memory(). [{total,15168232704}, {processes,14044779256}, {processes_used,14044755120}, {system,1123453448}, {atom,1057033}, {atom_used,1052587}, {binary,236381264}, {code,21790238}, {ets,391423744}] erlang:system_info(process_count). 1850309 ``` 7 procs per connection + 15 per session (7 + 2*15) * 50,000 = 1,850,000 procs 50k connections + 100k session require with this commit: 4.5 GB in RabbitMQ 3.x: 15 GB ## Future work 1. More efficient parser and serializer 2. TODO in mc_amqp: Do not store the parsed message on disk. 3. Implement both AMQP HTTP extension and AMQP management extension to allow AMQP clients to create RabbitMQ objects (queues, exchanges, ...).
2023-07-21 18:29:07 +08:00
Con = #consumer{last_msg_id = LastMsgId} = maps:get(Tag, CDels0),
CDels = update_consumer(Tag, LastMsgId, DelCntIncr, Con, CDels0),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
State#state{consumers = CDels}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
get_missing_deliveries(State, From, To, ConsumerTag) ->
%% find local server
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ConsumerKey = consumer_key(ConsumerTag, State),
?LOG_DEBUG("get_missing_deliveries for consumer '~s' from ~b to ~b",
2025-07-16 06:32:20 +08:00
[ConsumerTag, From, To]),
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
Cmd = {get_checked_out, ConsumerKey, lists:seq(From, To)},
ServerId = find_local_or_leader(State),
case ra:aux_command(ServerId, Cmd) of
{ok, Missing} ->
Missing;
{error, Error} ->
{protocol_error, internal_error, "Cannot query missing deliveries from ~tp: ~tp",
[ServerId, Error]};
{timeout, _} ->
{protocol_error, internal_error, "Cannot query missing deliveries from ~tp: timeout",
[ServerId]}
end.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
pick_server(#state{leader = undefined,
2020-07-10 21:25:17 +08:00
cfg = #cfg{servers = [N | _]}}) ->
%% TODO: pick random rather that first?
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
N;
pick_server(#state{leader = Leader}) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Leader.
% servers sorted by last known leader
sorted_servers(#state{leader = undefined,
cfg = #cfg{servers = Servers}}) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Servers;
sorted_servers(#state{leader = Leader,
cfg = #cfg{servers = Servers}}) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
[Leader | lists:delete(Leader, Servers)].
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
consumer_key(ConsumerTag, #state{consumers = Consumers}) ->
case Consumers of
#{ConsumerTag := #consumer{key = Key}} ->
Key;
_ ->
%% if no consumer found fall back to using the ConsumerId
consumer_id(ConsumerTag)
end.
consumer_id(ConsumerTag) when is_binary(ConsumerTag) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
{ConsumerTag, self()}.
send_command(Server, Correlation, Command, Priority,
#state{pending = Pending,
next_seq = Seq,
cfg = #cfg{soft_limit = SftLmt}} = State) ->
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
ok = case rabbit_fifo:is_return(Command) of
true ->
%% returns are sent to the aux machine for pre-evaluation
ra:cast_aux_command(Server, {Command, Seq, self()});
_ ->
ra:pipeline_command(Server, Command, Seq, Priority)
end,
State#state{pending = Pending#{Seq => {Correlation, Command}},
next_seq = Seq + 1,
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
slow = map_size(Pending) >= SftLmt}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
resend_command(ServerId, Correlation, Command,
#state{pending = Pending,
next_seq = Seq} = State) ->
ok = ra:pipeline_command(ServerId, Command, Seq),
State#state{pending = Pending#{Seq => {Correlation, Command}},
next_seq = Seq + 1}.
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
add_command(_, _, [], Acc) ->
Quorum queues (#1706) * Test queue.declare method with quorum type [#154472130] * Cosmetics [#154472130] * Start quorum queue Includes ra as a rabbit dependency [#154472152] * Update info and list operations to use quorum queues Basic implementation. Might need an update when more functionality is added to the quorum queues. [#154472152] * Stop quorum queue [#154472158] * Restart quorum queue [#154472164] * Introduce UId in ra config to support newer version of ra Improved ra stop [#154472158] * Put data inside VHost specific subdirs [#154472164] * Include ra in rabbit deps to support stop_app/start_app command [#154472164] * Stop quorum queues in `rabbit_amqqueue:stop/1` [#154472158] * Revert creation of fifo ets table inside rabbit Now supported by ra [#154472158] * Filter quorum queues [#154472158] * Test restart node with quorum queues [#154472164] * Publish to quorum queues [#154472174] * Use `ra:restart_node/1` [#154472164] * Wait for stats to be published when querying quorum queues [#154472174] * Test publish and queue length after restart [#154472174] * Consume messages from quorum queues with basic.get [#154472211] * Autoack messages from quorum queues on basic.get [#154472211] * Fix no_ack meaning no_ack = true is equivalent to autoack [#154472211] * Use data_dir as provided in the config If we modify the data_dir, ra is not able to delete the data when a queue is deleted [#154472158] * Remove unused code/variables [#154472158] * Subscribe to a quorum queue Supports auto-ack [#154472215] * Ack messages consumed from quorum queues [#154472221] * Nack messages consumed from quorum queues [#154804608] * Use delivery tag as consumer tag for basic.get in quorum queues [#154472221] * Support for publisher confirms in quorum queues [#154472198] * Integrate with ra_fifo_client * Clear queue state on queue.delete [#154472158] * Fix quorum nack [#154804608] * Test redelivery after nack [#154804608] * Nack without requeueing [#154472225] * Test multiple acks [#154804208] * Test multiple nacks [#154804314] * Configure dead letter exchange with queue declare [#155076661] * Use a per-vhost process to handle dead-lettering Needs to hold state for quorum queues [#155401802] * Implement dead-lettering on nack'ed messages [#154804620] * Use queue name as a resource on message delivery Fixes a previously introduced bug [#154804608] * Handle ra events on dead letter process [#155401802] * Pass empty queue states to queue delete Queue deletion on vhost deletion calls directly to rabbit_amqqueue. Queue states are not available, but we can provide an empty map as in deletion the states are only needed for cleanup. * Generate quorum queue stats and events Consumer delete events are still pending, as depend on basic.cancel (not implemented yet), ra terminating or ra detecting channel down [#154472241] * Ensure quorum mapping entries are available before metric emission [#154472241] * Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var [#154472152] * Use untracked enqueues when sending wihtout channel Updated several other calls missed during the quorum implementation * Revert "Configure data_dir, uses new RABBITMQ_QUORUM_BASE env var" This reverts commit f2261212410affecb238fcbd1fb451381aee4036. * Configure data_dir, uses new RABBITMQ_QUORUM_DIR based on mnesia dir [#154472152] * Fix get_quorum_state * Fix calculation of quorum pids * Move all quorum queues code to its own module [#154472241] * Return an error when declaring a quorum queue with an incompatible argument [#154521696] * Cleanup of quorum queue state after queue delete Also fixes some existing problems where the state wasn't properly stored [#155458625] * Revert Revert "Declare a quorum queue using the queue.declare method" * Remove duplicated state info [#154472241] * Start/stop multi-node quorum queue [#154472231] [#154472236] * Restart nodes in a multi-node quorum cluster [#154472238] * Test restart and leadership takeover on multiple nodes [#154472238] * Wait for leader down after deleting a quorum cluster It ensures an smooth delete-declare sequence without race conditions. The test included here detected the situation before the fix. [#154472236] * Populate quorum_mapping from mnesia when not available Ensures that leader nodes that don't have direct requests can get the mapping ra name -> queue name * Cosmetics * Do not emit core metrics if queue has just been deleted * Use rabbit_mnesia:is_process_alive Fixes bug introduced by cac9583e1bb2705be7f06c2ab7f416a75d11c875 [#154472231] * Only try to report stats if quorum process is alive * Implement cancel consumer callback Deletes metrics and sends consumer deleted event * Remove unnecessary trigger election call ra:restart_node has already been called during the recovery * Apply cancellation callback on node hosting the channel * Cosmetics * Read new fifo metrics which store directly total, ready and unack * Implement basic.cancel for quorum queues * Store leader in amqqueue record, report all in stats [#154472407] * Declare quorum queue in mnesia before starting the ra cluster Record needs to be stored first to update the leader on ra effects * Revert * Purge quorum queues [#154472182] * Improve use of untracked_enqueue Choose the persisted leader id instead of just using the id of the leader at point of creation. * Store quorum leader in the pid field of amqqueue record Same as mirrored queues, no real need for an additional field * Improve recovery When a ra node has never been started on a rabbit node ensure it doesn't fail but instead rebuilds the config and starts the node as a new node. Also fix issue when a quorum queue is declared when one of it's rabbit nodes are unavailable. [#157054606] * Cleanup core metrics after leader change [#157054473] * Return an error on sync_queue on quorum queues [#154472334] * Return an error on cancel_sync_queue on quorum queues [#154472337] * Fix basic_cancel and basic_consume return values Ensure the quorum queue state is always returned by these functions. * Restore arity of amqqeueu delete and purge functions. This avoids some breaking changes in the cli. * Fix bug returning consumers. * remove rogue debug log * Integrate ingress flow control with quorum queues [#157000583] * Configure commands soft limit [#157000583] * Support quorum pids on rabbit_mnesia:is_process_alive * Publish consumers metric for quorum queues * Whitelist quorum directory in is_virgin_node Allow the quorum directoy to exist without affecting the status of the Rabbit node. * Delete queue_metrics on leader change. Also run the become_leader handler in a separate process to avoid blocking. [#157424225] * Report cluster status in quorum queue infos. New per node status command. Related to [#157146500] * Remove quorum_mapping table As we can store the full queue name resource as the cluster id of the ra_fifo_client state we can avoid needed the quorum_mapping table. * Fix xref issue * Provide quorum members information in stats [#157146500] * fix unused variable * quorum queue multiple declare handling Extend rabbit_amqqueue:internal_declare/2 to indicate if the queue record was created or exisiting. From this we can then provide a code path that should handle concurrent queue declares of the same quorum queue. * Return an error when declaring exclusive/auto-delete quorum queue [#157472160] * Restore lost changes from 79c9bd201e1eac006a42bd162e7c86df96496629 * recover another part of commit * fixup cherry pick * Ra io/file metrics handler and stats publishing [#157193081] * Revert "Ra io/file metrics handler and stats publishing" This reverts commit 05d15c786540322583fc655709825db215b70952. * Do not issue confirms on node down for quorum queues. Only a ra_event should be used to issue positive confirms for a quorum queue. * Ra stats publishing [#157193081] * Pick consumer utilisation from ra data [#155402726] * Handle error when deleting a quorum queue and all nodes are already down This is in fact a successful deletion as all raft nodes are already 'stopped' [#158656366] * Return an error when declaring non-durable quorum queues [#158656454] * Rename dirty_query to committed_query * Delete stats on leader node [#158661152] * Give full list of nodes to fifo client * Handle timeout in quorum basic_get * Fix unused variable error * Handle timeout in basic get [#158656366] * Force GC after purge [#158789389] * Increase `ra:delete_cluster` timeout to 120s * Revert "Force GC after purge" This reverts commit 5c98bf22994eb39004760799d3a2c5041d16e9d4. * Add quorum member command [#157481599] * Delete quorum member command [#157481599] * Implement basic.recover for quorum queues [#157597411] * Change concumer utilisation to use the new ra_fifo table and api. * Set max quorum queue size limit Defaults to 7, can be configured per queue on queue.declare Nodes are selected randomly from the list of nodes, but the one that is executing the queue.declare command [#159338081] * remove potentially unrelated changes to rabbit_networking * Move ra_fifo to rabbit Copied ra_fifo to rabbit and renamed it rabbit_fifo. [#159338031] * rabbit_fifo tidy up * rabbit_fifo tidy up * rabbit_fifo: customer -> consumer rename * Move ra_fifo tests [#159338031] * Tweak quorum_queue defaults * quorum_queue test reliability * Optimise quorum_queue test suite. By only starting a rabbit cluster per group rather than test. [#160612638] * Renamings in line with ra API changes * rabbit_fifo fixes * Update with ra API changes Ra has consolidated and simplified it's api. These changes update to confirm to that. * Update rabbit_fifo with latest ra changes * Clean up out of date comment * Return map of states * Add test case for basic.get on an empty queue Before the previous patch, any subsequent basic.get would crash as the map of states had been replaced by a single state. * Clarify use of deliver tags on record_sent * Clean up queues after testcase * Remove erlang monitor of quorum queues in rabbit_channel The eol event can be used instead * Use macros to make clearer distinctions between quorum/classic queues Cosmetic only * Erase queue stats on 'eol' event * Update to follow Ra's cluster_id -> cluster_name rename. * Rename qourum-cluster-size To quorum-initial-group-size * Issue confirms on quorum queue eol Also avoid creating quorum queue session state on queue operation methods. * Only classic queues should be notified on channel down * Quorum queues do not support global qos Exit with protocol error of a basic.consume for a quorum queue is issued on a channel with global qos enabled. * unused variable name * Refactoring Strictly enfornce that channels do not monitor quorum queues. * Refactor foreach_per_queue in the channel. To make it call classic and quorum queues the same way. [#161314899] * rename function * Query classic and quorum queues separately during recovery as they should not be marked as stopped during failed vhost recovery. * Remove force_event_refresh function As the only user of this function, the management API no longer requires it. * fix errors * Remove created_at from amqqueue record [#161343680] * rabbit_fifo: support AMQP 1.0 consumer credit This change implements an alternative consumer credit mechanism similar to AMQP 1.0 link credit where the credit (prefetch) isn't automatically topped up as deliveries are settled and instead needs to be manually increased using a credit command. This is to be integrated with the AMQP 1.0 plugin. [#161256187] * Add basic.credit support for quorum queues. Added support for AMQP 1.0 transfer flow control. [#161256187] * Make quorum queue recover idempotent So that if a vhost crashes and runs the recover steps it doesn't fail because ra servers are still running. [#161343651] * Add tests for vhost deletion To ensure quorum queues are cleaned up on vhost removal. Also fix xref issue. [#161343673] * remove unused clause * always return latest value of queue * Add rabbitmq-queues scripts. Remove ra config from .bat scripts. * Return error if trying to get quorum status of a classic queue.
2018-10-29 17:47:29 +08:00
Acc;
add_command(Cid, settle, MsgIds, Acc) ->
[rabbit_fifo:make_settle(Cid, MsgIds) | Acc];
add_command(Cid, return, MsgIds, Acc) ->
[rabbit_fifo:make_return(Cid, MsgIds) | Acc];
add_command(Cid, discard, MsgIds, Acc) ->
[rabbit_fifo:make_discard(Cid, MsgIds) | Acc].
set_timer(QName, #state{leader = Leader0,
cfg = #cfg{servers = [Server | _]}} = State) ->
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
Leader = case Leader0 of
undefined -> Server;
_ ->
Leader0
end,
Ref = erlang:send_after(?TIMER_TIME, self(),
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
{'$gen_cast',
{queue_event, QName, {Leader, timeout}}}),
State#state{timer_state = Ref}.
cancel_timer(#state{timer_state = undefined} = State) ->
State;
cancel_timer(#state{timer_state = Ref} = State) ->
2023-01-24 05:26:14 +08:00
_ = erlang:cancel_timer(Ref, [{async, true}, {info, false}]),
State#state{timer_state = undefined}.
find_local_or_leader(#state{leader = Leader,
cfg = #cfg{servers = Servers}}) ->
case find_local(Servers) of
undefined ->
Leader;
ServerId ->
ServerId
end.
find_local([{_, N} = ServerId | _]) when N == node() ->
ServerId;
find_local([_ | Rem]) ->
find_local(Rem);
find_local([]) ->
undefined.
find_leader([]) ->
undefined;
find_leader([Server | Servers]) ->
case ra:members(Server, 500) of
{ok, _, Leader} -> Leader;
_ ->
find_leader(Servers)
end.
Stream Queue This is an aggregated commit of all changes related to the initial implementation of queue types and on top of that the stream queue type. The varios commit messages have simply been included mostly un-edited below. Make rabbit_amqqueue:not_found_or_absent_dirty/1 visible For use in the stream plugin. Use bigger retention policy on max-age test Set coordinator timeout to 30s Handle coordinator unavailable error Handle operator policies as maps when checking if is applicable Add is_policy_applicable/2 to classic queues Ignore restart commands if the stream has been deleted It could happen that after termination some of the monitors are still up and trigger writer/replica restarts Policy support on stream queues Remove subscription events on stream coordinator Ensure old leaders are removed from monitors Introduce delay when retrying a failed phase Note that this ensures monitor is setup, there was a bug where no monitor was really started when re-trying the same phase Restart replicas after leader election instead of relying on old monitors Use timer for stream coordinator retries Fix stream stats for members/online Multiple fixes for replica monitoring and restart Ensure pending commands are appended at the end and re-run Ensure phase is reset with the state Remove duplicates from replica list Restart current phase on state_enter Remove unused import Ensure rabbit is running when checking for stream quorum Restart replicas Add a close/1 function to queue types So that we can get a chance of cleaning up resources if needed. Stream queues close their osiris logs at this point. fix compiler errors stream-queue: take retention into account When calculating ready messages metrics. Add osiris to the list of rabbit deps Retry restart of replicas Do not restart replicas or leaders after receiving a delete cluster command Add more logging to the stream coordinator Monitor subscribed processes on the stream coordinator Memory breakdown for stream queues Update quorum queue event formatter rabbit_msg_record fixes Refactor channel confirms Remove old unconfirmed_messages module that was designed to handle multiple queue fan in logic including all ha mirrors etc. Replaced with simpler rabbit_confirms module that handles the fan out and leaves any queue specific logic (such as confirms from mirrors) to the queue type implemention. Also this module has a dedicated test module. Which is nice. Backward compatibility with 3.8.x events Supports mixed version cluster upgrades Match specification when stream queue already exists Max age retention for stream queues Stop all replicas before starting leader election stream: disallow global qos remove IS_CLASSIC|QUORUM macros Ensure only classic queues are notified on channel down This also removes the delivering_queues map in the channel state as it should not be needed for this and just cause additional unecessary accounting. Polish AMQP 1.0/0.9.1 properties conversion Support byte in application properties, handle 1-bit representation for booleans. Use binary in header for long AMQP 1.0 ID Fix AMQP 1.0 to 0.9.1 conversion Fix test due to incorrect type Convert timestamp application properties to/from seconds AMQP 1.0 uses milliseconds for timestamp and AMQP 0.9.1 uses seconds, so conversion needed. Dialyzer fixes Handle all message-id types AMQP 1.0 is more liberal in it's allowed types of message-id and correlation-id - this adds headers to describe the type of the data in the message_id / correlation_id properties and also handles the case where the data cannot fit by again using headers. Resize stream coordinator cluster when broker configuration changes convert timestamp to and fro seconds user_id should be a binary message annotations keys need to be symbols stream-queue: default exchange and routing key As these won't be present for data written using the rabbitmq-stream plugin. Add exchange, routing key as message annotations To the AMQP 1.0 formatted data to enable roundtrip. Add osiris logging module config And update logging config test suite. Restart election when start of new leader fails The node might have just gone down so we need to try another one Only aux keeps track of phase now, as it might change if the leader election fails Stream coordinator refactor - all state is kept on the ra machine Ensure any ra cluster not a qq is not cleaned up Fixes to recovery and monitoring Add AMQP 1.0 common to dependencies Add rabbit_msg_record module To handle conversions into internal stream storage format. Use rabbitmq-common stream-queue branch Use SSH for osiris dependency Stream coordinator: delete replica Stream coordinator: add replica Stream coordinator: leader failover Stream coordinator: declare and delete Test consuming from a random offset Previous offsets should not be delivered to consumers Consume from stream replicas and multiple test fixes Use max-length-bytes and add new max-segment-size Use SSH for osiris dependency Basic cancel for stream queues Publish stream queues and settle/reject/requeue refactor Consume from stream queues Fix recovery Publish stream messages Add/delete stream replicas Use safe queue names Set retention policy for stream queues Required by the ctl command [#171207092] Stream queue delete queue fix missing callback impl Stream queue declare Queue type abstraction And use the implementing module as the value of the amqqueue record `type` field. This will allow for easy dispatch to the queue type implementation. Queue type abstraction Move queue declare into rabbit_queue_type Move queue delete into queue type implementation Queue type: dequeue/basic_get Move info inside queue type abstraction Move policy change into queue type interface Add purge to queue type Add recovery to the queue type interface Rename amqqueue quorum_nodes field To a more generic an extensible opaque queue type specific map. Fix tests and handle classic API response Fix HA queue confirm bug All mirrors need to be present as queue names. This introduces context linking allowing additional queue refs to be linked to a single "master" queue ref contining the actual queue context. Fix issue with events of deleted queues Also update queue type smoke test to use a cluster by default. correct default value of amqqueue getter Move classic queues further inside queue type interface why [TrackerId] Dialyzer fixes
2020-09-29 18:43:24 +08:00
qref({Ref, _}) -> Ref;
qref(Ref) -> Ref.
-spec cluster_name(state()) ->
atom().
cluster_name(#state{cfg = #cfg{servers = [{Name, _Node} | _]}}) ->
Name.
Quorum queues v4 (#10637) This commit contains the following new quorum queue features: * Fair share high/low priorities * SAC consumers honour consumer priorities * Credited consumer refactoring to meet AMQP requirements. * Use checkpoints feature to reduce memory use for queues with long backlogs * Consumer cancel option that immediately removes consumer and returns all pending messages. * More compact commands of the most common commands such as enqueue, settle and credit * Correctly track the delivery-count to be compatible with the AMQP spec * Support the "modified" AMQP 1.0 outcome better. Commits: * Quorum queues v4 scaffolding. Create the new version but not including any changes yet. QQ: force delete followers after leader has terminated. Also try a longer sleep for mqtt_shared_SUITE so that the delete operation stands a chance to time out and move on to the forced deletion stage. In some mixed machine version scenarios some followers will never apply the poison pill command so we may as well force delete them just in case. QQ: skip test in amqp_client that cannot pass with mixed machine versions QQ: remove dead code Code relating to prior machine versions and state conversions. rabbit_fifo_prop_SUITE fixes * QQ: add v4 ff and new more compact enqueue command. Also update rabbit_fifo_* suites to test more relevant code versions where applicable. QQ: always use the updated credit mode format QQv4: use more compact consumer reference in settle, credit, return This introudces a new type: consumer_key() which is either the consumer_id or the raft index the checkout was processed at. If the consumer is using one of the updated credit spec formats rabbit_fifo will use the raft index as the primary key for the consumer such that the rabbit fifo client can then use the more space efficient integer index instead of the full consumer id in subsequent commands. There is compatibility code to still accept the consumer id in settle, return, discard and credit commands but this is slighlyt slower and of course less space efficient. The old form will be used in cases where the fifo client may have already remove the local consumer state (as happens after a cancel). Lots of test refactorings of the rabbit_fifo_SUITE to begin to use the new forms. * More test refactoring and new API fixes rabbit_fifo_prop_SUITE refactoring and other fixes. * First pass SAC consumer priority implementation. Single active consumers will be activated if they have a higher priority than the currently active consumer. if the currently active consumer has pending messages, no further messages will be assigned to the consumer and the activation of the new consumer will happen once all pending messages are settled. This is to ensure processing order. Consumers with the same priority will internally be ordered to favour those with credit then those that attached first. QQ: add SAC consumer priority integration tests QQ: add check for ff in tests * QQ: add new consumer cancel option: 'remove' This option immediately removes and returns all messages for a consumer instead of the softer 'cancel' option which keeps the consumer around until all pending messages have been either settled or returned. This involves a change to the rabbit_queue_type:cancel/5 API to rabbit_queue_type:cancel/3. * QQ: capture checked out time for each consumer message. This will form the basis for queue initiated consumer timeouts. * QQ: Refactor to use the new ra_machine:handle_aux/5 API Instead of the old ra_machine:handle_aux/6 callback. * QQ hi/lo priority queue * QQ: Avoid using mc:size/1 inside rabbit_fifo As we dont want to depend on external functions for things that may change the state of the queue. * QQ bug fix: Maintain order when returning multiple Prior to this commit, quorum queues requeued messages in an undefined order, which is wrong. This commit fixes this bug and requeues messages always in the order as nacked / rejected / released by the client. We ensure that order of requeues is deterministic from the client's point of view and doesn't depend on whether the quorum queue soft limit was exceeded temporarily. So, even when rabbit_fifo_client batches requeues, the order as nacked by the client is still maintained. * Simplify * Add rabbit_quorum_queue:file_handle* functions back. For backwards compat. * dialyzer fix * dynamic_qq_SUITE: avoid mixed versions failure. * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * Use QQ consumer removal when AMQP client detaches This enables us to unskip some AMQP tests. * Use AMQP address v2 in fsharp-tests * QQ: track number of requeues for message. To be able to calculate the correct value for the AMQP delivery_count header we need to be able to distinguish between messages that were "released" or returned in QQ speak and those that were returned due to errors such as channel termination. This commit implement such tracking as well as the calculation of a new mc annotations `delivery_count` that AMQP makes use of to set the header value accordingly. * rabbit_fifo: Use Ra checkpoints * quorum queues: Use a custom interval for checkpoints * rabbit_fifo_SUITE: List actual effects in ?ASSERT_EFF failure * QQ: Checkpoints modifications * fixes * QQ: emit release cursors on tick for followers and leaders else followers could end up holding on to segments a bit longer after traffic stops. * Support draining a QQ SAC waiting consumer By issuing drain=true, the client says "either send a transfer or a flow frame". Since there are no messages to send to an inactive consumer, the sending queue should advance the delivery-count consuming all link-credit and send a credit_reply with drain=true to the session proc which causes the session proc to send a flow frame to the client. * Extract applying #credit{} cmd into 2 functions This commit is only refactoring and doesn't change any behaviour. * Fix default priority level Prior to this commit, when a message didn't have a priority level set, it got enqueued as high prio. This is wrong because the default priority is 4 and "for example, if 2 distinct priorities are implemented, then levels 0 to 4 are equivalent, and levels 5 to 9 are equivalent and levels 4 and 5 are distinct." Hence, by default a message without priority set, must be enqueued as low prio. * bazel run gazelle * Avoid deprecated time unit * Fix aux_test * Delete dead code * Fix rabbit_fifo_q:get_lowest_index/1 * Delete unused normalize functions * Generate less garbage * Add integration test for QQ SAC with consumer priority * Improve readability * Change modified outcome behaviour With the new quorum queue v4 improvements where a requeue counter was added in addition to the quorum queue delivery counter, the following sentence from https://github.com/rabbitmq/rabbitmq-server/pull/6292#issue-1431275848 doesn't apply anymore: > Also the case where delivery_failed=false|undefined requires the release of the > message without incrementing the delivery_count. Again this is not something > that our queues are able to do so again we have to reject without requeue. Therefore, we simplify the modified outcome behaviour: RabbitMQ will from now on only discard the message if the modified's undeliverable-here field is true. * Introduce single feature flag rabbitmq_4.0.0 ## What? Merge all feature flags introduced in RabbitMQ 4.0.0 into a single feature flag called rabbitmq_4.0.0. ## Why? 1. This fixes the crash in https://github.com/rabbitmq/rabbitmq-server/pull/10637#discussion_r1681002352 2. It's better user experience. * QQ: expose priority metrics in UI * Enable skipped test after rebasing onto main * QQ: add new command "modify" to better handle AMQP modified outcomes. This new command can be used to annotate returned or rejected messages. This commit also retains the delivery-count across dead letter boundaries such that the AMQP header delivery-count field can now include _all_ failed deliver attempts since the message was originally received. Internally the quorum queue has moved it's delivery_count header to only track the AMQP protocol delivery attempts and now introduces a new acquired_count to track all message acquisitions by consumers. * Type tweaks and naming * Add test for modified outcome with classic queue * Add test routing on message-annotations in modified outcome * Skip tests in mixed version tests Skip tests in mixed version tests because feature flag rabbitmq_4.0.0 is needed for the new #modify{} Ra command being sent to quorum queues. --------- Co-authored-by: David Ansari <david.ansari@gmx.de> Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
2024-08-08 15:48:27 +08:00
send_pending(Cid, #state{unsent_commands = Unsent} = State0) ->
Commands = case Unsent of
#{Cid := {Settled, Returns, Discards}} ->
add_command(Cid, settle, Settled,
add_command(Cid, return, Returns,
add_command(Cid, discard,
Discards, [])));
_ ->
[]
end,
ServerId = pick_server(State0),
%% send all the settlements, discards and returns
State1 = lists:foldl(fun (C, S0) ->
send_command(ServerId, undefined, C,
normal, S0)
end, State0, Commands),
State1#state{unsent_commands = maps:remove(Cid, Unsent)}.
now_ms() ->
erlang:system_time(millisecond).