Deprecate existing metadata query APIs in favor of new
ones that include standby hosts as well as partition
information.
Closes: #7960
Implements: KIP-535
Co-authored-by: Navinder Pal Singh Brar <navinder_brar@yahoo.com>
Reviewed-by: John Roesler <vvcephei@apache.org>
Do not initialize `JmxTool` by default when running console consumer. In order to support this, we remove `has_partitions_assigned` and its only usage in an assertion inside `ProduceConsumeValidateTest`, which did not seem to contribute much to the validation.
Reviewers: David Arthur <mumrah@gmail.com>, Jason Gustafson <jason@confluent.io>
In some system tests a Streams app is started and then prints a message to stdout, which the system test waits for to confirm the node has successfully been brought up. It then greps for certain log messages in a retriable loop.
But waiting on the Streams app to start/print to stdout does not mean the log file has been created yet, so the grep may return an error. Although this occurs in a retriable loop it is assumed that grep will not fail, and the result is piped to wc and then blindly converted to an int in the python function, which fails since the error message is a string (throws ValueError)
We should catch the ValueError and return a 0 so it can try again rather than immediately crash
Reviewers: Bill Bejeck <bbejeck@gmail.com>, John Roesler <vvcephei@users.noreply.github.com>, Guozhang Wang <wangguoz@gmail.com>
The upgrade system test correctly rolls by upgrading the broker and
leaving the IBP, and then rolling again with the latest IBP version.
Unfortunately, this is not sufficient to pick up many problems in our IBP
gating as we charge through the rolls and after the second roll all of
the brokers will rejoin the ISR and the test will be treated as a
success.
This test adds two new checks:
1. We wait for the ISR to stabilize for all partitions. This is best
practice during rolls, and is enough to tell us if a broker hasn't
rejoined after each roll.
2. We check the broker logs for some common protocol errors. This is a
fail safe as it's possible for the test to be successful even if some
protocols are incompatible and the ISR is rejoined.
Reviewers: Nikhil Bhatia <nikhil@confluent.io>, Jason Gustafson <jason@confluent.io>
The --enable-autocommit argument is a flag. It does not take a parameter. This was broken in #7724.
Reviewers: Ismael Juma <ismael@juma.me.uk>, Manikumar Reddy <manikumar.reddy@gmail.com>
Two tests using 50k replicas on 8 brokers:
* Do a rolling restart with clean shutdown, delete topics
* Run produce bench and consumer bench on a subset of topics
Reviewed-By: David Jacot <djacot@confluent.io>, Vikas Singh <vikas@confluent.io>, Jason Gustafson <jason@confluent.io>
This patch adds a basic downgrade system test. It verifies that producing and consuming continues to work before and after the downgrade.
Reviewers: Ismael Juma <ismael@juma.me.uk>, David Arthur <mumrah@gmail.com>
Recently, system tests test_rebalance_[simple|complex] failed
repeatedly with a verfication error. The cause was most probably
the missing clean-up of a state directory of one of the processors.
A node is cleaned up when a service on that node is started and when
a test is torn down.
If the clean-up flag clean_node_enabled of a EOS Streams service is
unset, the clean-up of the node is skipped.
The clean-up flag of processor1 in the EOS tests should stay set before
its first start, so that the node is cleaned before the service is started.
Afterwards for the multiple restarts of processor1 the cleans-up flag should
be unset to re-use the local state.
After the multiple restarts are done, the clean-up flag of processor1 should
again be set to trigger node clean-up during the test teardown.
A dirty node can lead to test failures when tests from Streams EOS tests are
scheduled on the same node, because the state store would not start empty
since it reads the local state that was not cleaned up.
Reviewers: Matthias J. Sax <mjsax@apache.org>, Andrew Choi <andchoi@linkedin.com>, Bill Bejeck <bbejeck@gmail.com>
* Add rate limiting to tc
* Feedback from PR
* Add a sanity test for tc
* Add iperf to vagrant scripts
* Dynamically determine the network interface
* Add some temp code for testing on AWS
* Temp: use hostname instead of external IP
* Temp: more AWS debugging
* More AWS WIP
* More AWS temp
* Lower latency some
* AWS wip
* Trying this again now that ping should work
* Add cluster decorator to tests
* Fix broken import
* Fix device name
* Fix decorator arg
* Remove errant import
* Increase timeouts
* Fix tbf command, relax assertion on latency test
* Fix log line
* Final bit of cleanup
* Newline
* Revert Trogdor retry count
* PR feedback
* More PR feedback
* Feedback from PR
* Remove unused argument
When Trogdor wants to clear all the faults injected to Kibosh, it sends the empty JSON object {}. However, Kibosh expects {"faults":[]} instead. Kibosh should handle the empty JSON object, since that's consistent with how Trogdor handles empty JSON fields in general (if they're empty, they can be omitted). We should also have a test for this.
Reviewers: David Arthur <mumrah@gmail.com>, Stanislav Kozlovski <stanislav_kozlovski@outlook.com>
Instead of caching the checkpoint map during StandbyTask
initialization, use the latest checkpoints (which would have
been updated during suspend).
Reviewers: Bill Bejeck <bill@confluent.io>
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Vikas Singh <vikas@confluent.io>, Jason Gustafson <jason@confluent.io>
Closes#7477 from cmccabe/KAFKA-8984
The consumer's `committed` API does not return an entry in the response map for a requested partition if there is no committed offset. The transactional message copier, which is used in the transaction system test, did not account for this. If the first transaction attempted by the copier was randomly aborted, then we would not seek to the beginning as expected, which means we would fail to copy some of the records.
This patch fixes the problem by iterating over the assignment rather than the result of `committed` when resetting offsets. It also adds enables additional logging in the transaction message copier service to make finding problems easier in the future.
Author: Jason Gustafson <jason@confluent.io>
Reviewers: Manikumar Reddy <manikumar.reddy@gmail.com>
Closes#7653 from hachikuji/fix-transaction-system-test
With KIP-444 the metrics definitions are refactored. Thus, Streams' SimpleBenchmark needs to be updated to correctly access the refactored metrics.
Reviewers: Guozhang Wang <wangguoz@gmail.com>, Matthias J. Sax <mjsax@apache.org>, Bill Bejeck <bbejeck@gmail.com>
Inside onLeavePrepare we would look into the assignment and try to revoke the owned tasks and notify users via RebalanceListener#onPartitionsRevoked, and then clear the assignment.
However, the subscription's assignment is already cleared in this.subscriptions.unsubscribe(); which means user's rebalance listener would never be triggered. In other words, from consumer client's pov nothing is owned after unsubscribe, but from the user caller's pov the partitions are not revoked yet. For callers like Kafka Streams which rely on the rebalance listener to maintain their internal state, this leads to inconsistent state management and failure cases.
Before KIP-429 this issue is hidden away since every time the consumer re-joins the group later, it would still revoke everything anyways regardless of the passed-in parameters of the rebalance listener; with KIP-429 this is easier to reproduce now.
Our fixes are following:
• Inside unsubscribe, first do onLeavePrepare / maybeLeaveGroup and then subscription.unsubscribe. This we we are guaranteed that the streams' tasks are all closed as revoked by then.
• [Optimization] If the generation is reset due to fatal error from join / hb response etc, then we know that all partitions are lost, and we should not trigger onPartitionRevoked, but instead just onPartitionsLost inside onLeavePrepare. This is because we don't want to commit for lost tracks during rebalance which is doomed to fail as we don't have any generation info.
Reviewers: Matthias J. Sax <matthias@confluent.io>, A. Sophie Blee-Goldman <sophie@confluent.io>, Bill Bejeck <bill@confluent.io>, Guozhang Wang <guozhang@confluent.io>
MM2 added a few connector classes in Connect's classpath and given that the assertion in the Connect REST system tests need to be adjusted to account for these additions.
This fix makes sure that the loaded Connect plugins are a superset of the expected by the test connectors.
Testing: The change is straightforward. The fix was tested with local system test runs.
Author: Konstantine Karantasis <konstantine@confluent.io>
Reviewers: Manikumar Reddy <manikumar.reddy@gmail.com>
Closes#7578 from kkonstantine/minor-fix-connect-test-after-mm2-classes
This system test was marked @Ignore around a year and a half ago pending the version probing work, but never turned on again.
These days, it is made redundant by the suite of system tests in streams_upgrade_test, which cover rolling upgrades (including version probing and metadata change).
Reviewers: Boyang Chen <boyang@confluent.io>, Bill Bejeck <bbejeck@gmail.com>
All four flavors of the repartition/optimization tests have been reported as flaky and failed in one place or another:
* RepartitionOptimizingIntegrationTest.shouldSendCorrectRecords_OPTIMIZED
* RepartitionOptimizingIntegrationTest.shouldSendCorrectRecords_NO_OPTIMIZATION
* RepartitionWithMergeOptimizingIntegrationTest.shouldSendCorrectRecords_OPTIMIZED
* RepartitionWithMergeOptimizingIntegrationTest.shouldSendCorrectRecords_NO_OPTIMIZATION
They're pretty similar so it makes sense to knock them all out at once. This PR does three things:
* Switch to in-memory stores wherever possible
* Name all operators and update the Topology accordingly (not really a flaky test fix, but had to update the topology names anyway because of the IM stores so figured might as well)
* Port to TopologyTestDriver -- this is the "real" fix, should make a big difference as these repartition tests required multiple roundtrips with the Kafka cluster (while using only the default timeout)
Reviewers: Bill Bejeck <bill@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Key improvements with this PR:
* tasks will remain available for IQ during a rebalance (but not during restore)
* continue restoring and processing standby tasks during a rebalance
* continue processing active tasks during rebalance until the RecordQueue is empty*
* only revoked tasks must suspended/closed
* StreamsPartitionAssignor tries to return tasks to their previous consumers within a client
* but do not try to commit, for now (pending KAFKA-7312)
Reviewers: John Roesler <john@confluent.io>, Boyang Chen <boyang@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Small follow-up to trunk PR #7423
While debugging the 2.3 VP PR we realized we should remove the leader-tracking from the VP system test altogether. We'd already merged the corresponding trunk PR so I made a quick new PR for trunk (also fixes a missed version bump in one of the log messages)
Reviewers: Guozhang Wang <wangguoz@gmail.com>
Implemented KIP-507 to secure the internal Connect REST endpoints that are only for intra-cluster communication. A new V2 of the Connect subprotocol enables this feature, where the leader generates a new session key, shares it with the other workers via the configuration topic, and workers send and validate requests to these internal endpoints using the shared key.
Currently the internal `POST /connectors/<connector>/tasks` endpoint is the only one that is secured.
This change adds unit tests and makes some small alterations to system tests to target the new `sessioned` Connect subprotocol. A new integration test ensures that the endpoint is actually secured (i.e., requests with missing/invalid signatures are rejected with a 400 BAD RESPONSE status).
Author: Chris Egerton <chrise@confluent.io>
Reviewed: Konstantine Karantasis <konstantine@confluent.io>, Randall Hauch <rhauch@gmail.com>
Instead of sending the leader's version and having older members try to blindly upgrade.
The only other real change here is that we will also set the VERSION_PROBING error code and return early from onAssignment when we are upgrading our used subscription version (not just downgrading it) since this implies the whole group has finished the rolling upgrade and all members should rejoin with the new subscription version.
Also piggy-backing on a fix for a potentially dangerous edge case, where every thread of an instance is assigned the same set of active tasks.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This change adds a command line option to the `ducker-ak up' command to enable exposing ports from docker containers. The exposed ports will be mapped to the ephemeral ports on the host. The option is called `expose-ports' and can take either a single value (like 5005) or a range (like 5005-5009). This port will then exposed from each docker container that ducker-ak sets up.
Reviewers: Colin P. McCabe <cmccabe@apache.org>, José Armando García Sancio <jsancio@users.noreply.github.com>
* Leader instance uses dictionary encoding on the wire to send topic partitions
* Topic names (most expensive component) are mapped to an integer using the dictionary
* Follower instances receive the dictionary, decode topic names back
* Purely an on-the-wire optimization, no in-memory structures changed
* Test case added for version 5 AssignmentInfo
Reviewers: Guozhang Wang <wangguoz@gmail.com>
Fix a bug where ClientCompatibilityFeaturesTest fails when running multiple iterations.
Also, fix a typo in tests/docker/Dockerfile.
Reviewers: Ismael Juma <ismael@juma.me.uk>
As part of commit 4d1ee26a13 streams
version 2.3.0 test jar was added, but there was a simple typo in the
path that specified the version.
`ducker-ak up` was failing because of that. Fixed that.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This adds a basic system test that enables rack-aware brokers with the rack-aware replica selector for fetch from followers (KIP-392). The test asserts that the follower was read from at least once and that all the messages that were produced were successfully consumed.
Reviewers: Jason Gustafson <jason@confluent.io>
Corrected the AbstractHerder to correctly identify task configs that contain variables for externalized secrets. The original method incorrectly used `matcher.matches()` instead of `matcher.find()`. The former method expects the entire string to match the regex, whereas the second one can find a pattern anywhere within the input string (which fits this use case more correctly).
Added unit tests to cover various cases of a config with externalized secrets, and updated system tests to cover case where config value contains additional characters besides secret that requires regex pattern to be found anywhere in the string (as opposed to complete match).
Author: Arjun Satish <arjun@confluent.io>
Reviewer: Randall Hauch <rhauch@gmail.com>
To avoid transient system test failures, tolerate a small amount of data loss due to truncation in upgrade system tests using older message format prior to KIP-101, where data loss was possible.
Reviewers: Ismael Juma <ismael@juma.me.uk>
ZooKeeper 3.5.5 is the first stable release in the 3.5.x series. The key new feature
in is TLS support, but there are a few more noteworthy features:
* Dynamic reconfiguration
* Local sessions
* New node types: Container, TTL
* Ability to remove watchers
* Multi-threaded commit processor
* Upgraded to Netty 4.1
See the release notes for more detail:
https://zookeeper.apache.org/doc/r3.5.5/releasenotes.html
In addition to the version bump, we:
* Add `commons-cli` dependency as it's required by `ZooKeeperMain`, but specified as
`provided` in their pom.
* Remove unnecessary `ZooKeeperMainWrapper`, the bug it worked around was fixed
upstream a long time ago.
* Ignore non zero exit in one system test invocation of `ZooKeeperMain`.
`ZooKeeperMainWrapper` always returned `0` and `ZooKeeperService.query` relies
on that for correct behavior.
Reviewers: Jason Gustafson <jason@confluent.io>
ZkUtils was removed so we don't need this anymore.
Also:
* Fix ZkSecurityMigrator and ReplicaManagerTest not to
reference ZkClient classes.
* Remove references to zkclient in various `log4j.properties`
and `import-control.xml`.
Reviewers: Manikumar Reddy <manikumar.reddy@gmail.com>, Stanislav Kozlovski <stanislav_kozlovski@outlook.com>
Follow on to #6731, this PR adds broker-side support for [KIP-392](https://cwiki.apache.org/confluence/display/KAFKA/KIP-392%3A+Allow+consumers+to+fetch+from+closest+replica) (fetch from followers).
Changes:
* All brokers will handle FetchRequest regardless of leadership
* Leaders can compute a preferred replica to return to the client
* New ReplicaSelector interface for determining the preferred replica
* Incremental fetches will include partitions with no records if the preferred replica has been computed
* Adds new JMX to expose the current preferred read replica of a partition in the consumer
Two new conditions were added for completing a delayed fetch. They both relate to communicating the high watermark to followers without waiting for a timeout:
* For regular fetches, if the high watermark changes within a single fetch request
* For incremental fetch sessions, if the follower's high watermark is lower than the leader
A new JMX attribute `preferred-read-replica` was added to the `kafka.consumer:type=consumer-fetch-manager-metrics,client-id=some-consumer,topic=my-topic,partition=0` object. This was added to support the new system test which verifies that the fetch from follower behavior works end-to-end. This attribute could also be useful in the future when debugging problems with the consumer.
Reviewers: José Armando García Sancio <jsancio@users.noreply.github.com>, Jun Rao <junrao@gmail.com>, Jason Gustafson <jason@confluent.io>
Connect tests were using String version for KafkaService instead of the expected KafkaVersion object. This broke due to recent changes to KafkaVersion. It turns out that the tests with String version were running compatibility tests against `dev` brokers rather than the older broker versions they were expecting to run against. When version was fixed, tests using 0.9.0.1 brokers started failing since new clients are not compatible with 0.9.0.1 brokers. So this PR fixes version parameter and removes the two tests against 0.9.0.1 brokers.
Reviewers: Ismael Juma <ismael@juma.me.uk>, Rajini Sivaram <rajinisivaram@googlemail.com>
This adds a new Trogdor fault spec for inducing network latency on a network device for system testing. It operates very similarly to the existing network partition spec by executing the `tc` linux utility.
This PR fixes a bug in static group membership. Previously we limit the `member.id` replacement in JoinGroup to only cases when the group is in Stable. This is error-prone and could potentially allow duplicate consumers reading from the same topic. For example, imagine a case where two unknown members join in the `PrepareRebalance` stage at the same time.
The PR fixes the following things:
1. Replace `member.id` at any time we see a known static member rejoins group with unknown member.id
2. Immediately fence any ongoing join/sync group callback to early terminate the duplicate member.
3. Clearly handle Dead/Empty cases as exceptional.
4. Return old leader id upon static member leader rejoin to avoid trivial member assignment being triggered.
Reviewers: Guozhang Wang <wangguoz@gmail.com>, Jason Gustafson <jason@confluent.io>
We've seen `ReplicaVerificationToolTest.test_replica_lags` fail occasionally due to errors such as the following:
```
RemoteCommandError: ubuntuworker7: Command 'kill -15 2896' returned non-zero exit status 1. Remote error message: bash: line 0: kill: (2896) - No such process
```
The problem seems to be a shutdown race condition when using `max_messages` with the producer. The process may already be gone which will cause the signal to fail.
Author: Jason Gustafson <jason@confluent.io>
Reviewers: Gwen Shapira
Closes#6906 from hachikuji/fix-failing-replicat-verification-test
We see the upgrade test failing from time to time. I looked into it and found that the root cause is basically that the test throughput can be too high for the 0.9 producer to make progress. Eventually it reaches a point where it has a huge backlog of timed out requests in the accumulator which all have to be expired. We see a long run of messages like this in the output:
```
{"exception":"class org.apache.kafka.common.errors.TimeoutException","time_ms":1559907386132,"name":"producer_send_error","topic":"test_topic","message":"Batch Expired","class":"class org.apache.kafka.tools.VerifiableProducer","value":"335160","key":null}
{"exception":"class org.apache.kafka.common.errors.TimeoutException","time_ms":1559907386132,"name":"producer_send_error","topic":"test_topic","message":"Batch Expired","class":"class org.apache.kafka.tools.VerifiableProducer","value":"335163","key":null}
{"exception":"class org.apache.kafka.common.errors.TimeoutException","time_ms":1559907386133,"name":"producer_send_error","topic":"test_topic","message":"Batch Expired","class":"class org.apache.kafka.tools.VerifiableProducer","value":"335166","key":null}
{"exception":"class org.apache.kafka.common.errors.TimeoutException","time_ms":1559907386133,"name":"producer_send_error","topic":"test_topic","message":"Batch Expired","class":"class org.apache.kafka.tools.VerifiableProducer","value":"335169","key":null}
```
This can continue for a long time (I have observed up to 1 min) and prevents the producer from successfully writing any new data. While it is busy expiring the batches, no data is getting delivered to the consumer, which causes it to eventually raise a timeout.
```
kafka.consumer.ConsumerTimeoutException
at kafka.consumer.NewShinyConsumer.receive(BaseConsumer.scala:50)
at kafka.tools.ConsoleConsumer$.process(ConsoleConsumer.scala:109)
at kafka.tools.ConsoleConsumer$.run(ConsoleConsumer.scala:69)
at kafka.tools.ConsoleConsumer$.main(ConsoleConsumer.scala:47)
at kafka.tools.ConsoleConsumer.main(ConsoleConsumer.scala)
```
The fix here is to reduce the throughput, which seems reasonable since the purpose of the test is to verify the upgrade, which does not demand heavy load. Note that I investigated several failing instances of this test going back to 1.0 and saw a similar pattern, so there does not appear to be a regression.
Author: Jason Gustafson <jason@confluent.io>
Reviewers: Gwen Shapira
Closes#6907 from hachikuji/lower-throughput-for-upgrade-test
As title suggested, we boost 3 stream instances stream job with one minute session timeout, and once the group is stable, doing couple of rolling bounces for the entire cluster. Every rejoin based on restart should have no generation bump on the client side.
Reviewers: Guozhang Wang <wangguoz@gmail.com>, Bill Bejeck <bbejeck@gmail.com>
Reviewers: Bill Bejeck <bill@confluent.io>, Boyang Chen <boyang@confluent.io>, Bruno Cadonna <bruno@confluent.io>, Guozhang Wang <guozhang@confuent.io>
These are important to ensure we don't break compatibility.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Gwen Shapira
Closes#6794 from ijuma/update-version-compat-tests
For static members join/rejoin, we encode the current timestamp in the new member.id. The format looks like group.instance.id-timestamp.
During consumer/broker interaction logic (Join, Sync, Heartbeat, Commit), we shall check the whether group.instance.id is known on group. If yes, we shall match the member.id stored on static membership map with the request member.id. If mismatching, this indicates a conflict consumer has used same group.instance.id, and it will receive a fatal exception to shut down.
Right now the only missing part is the system test. Will work on it offline while getting the major logic changes reviewed.
Reviewers: Ryanne Dolan <ryannedolan@gmail.com>, Jason Gustafson <jason@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Corrects the system tests to check for either a 404 or a 409 error and sleeping until the Connect REST API becomes available. This corrects a previous change to how REST extensions are initialized (#6651), which added the ability of Connect throwing a 404 if the resources are not yet started. The integration tests were already looking for 409.
Author: Magesh Nandakumar <magesh.n.kumar@gmail.com>
Reviewer: Randall Hauch <rhauch@gmail.com>
This is the first diff for the implementation of JoinGroup logic for static membership. The goal of this diff contains:
* Add group.instance.id to be unique identifier for consumer instances, provided by end user;
Modify group coordinator to accept JoinGroupRequest with/without static membership, refactor the logic for readability and code reusability.
* Add client side support for incorporating static membership changes, including new config for group.instance.id, apply stream thread client id by default, and new join group exception handling.
* Increase max session timeout to 30 min for more user flexibility if they are inclined to tolerate partial unavailability than burdening rebalance.
* Unit tests for each module changes, especially on the group coordinator logic. Crossing the possibilities like:
6.1 Dynamic/Static member
6.2 Known/Unknown member id
6.3 Group stable/unstable
6.4 Leader/Follower
The rest of the 345 change will be broken down to 4 separate diffs:
* Avoid kicking out members through rebalance.timeout, only do the kick out through session timeout.
* Changes around LeaveGroup logic, including version bumping, broker logic, client logic, etc.
* Admin client changes to add ability to batch remove static members
* Deprecate group.initial.rebalance.delay
Reviewers: Liquan Pei <liquanpei@gmail.com>, Stanislav Kozlovski <familyguyuser192@windowslive.com>, Jason Gustafson <jason@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
As titled, this PR changed the default reset policy to latest accidentally for system tests, which in fact was earliest.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
Verified that the https links work.
I didn't update the license header in this PR since that touches
so many files. Will file a separate one for that.
Reviewers: Manikumar Reddy <manikumar.reddy@gmail.com>
This patch adds a TimeIntervalTransactionsGenerator class which enables the Trogdor ProduceBench worker to commit transactions based on a configurable millisecond time interval.
Also, we now handle 409 create task responses in the coordinator command-line client by printing a more informative message
Reviewers: Colin P. McCabe <cmccabe@apache.org>
`kafka.list_topics(...)` should not require a topic parameter
Author: Brian Bushree <bbushree@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#6367 from brianbushree/list-topics-no-topic
* add a normal windowed suppress with short windows and a short grace
period
* improve the smoke test so that it actually verifies the intended
conditions
See https://issues.apache.org/jira/browse/KAFKA-7944
Reviewers: Bill Bejeck <bill@confluent.io>, Guozhang Wang <guozhang@confluent.io>
For consumers using SSL, this timeout includes the time to create and copy keystores and truststores and sometime it takes longer than 10s to complete the security setup before starting the consumer process.
Reviewers: Ismael Juma <ismael@juma.me.uk>
The StreamsBrokerBounceTest.test_broker_type_bounce experienced what looked like a transient failure. After looking over this test and failure, it seems like it is vulnerable to timing error that streams will start before the kafka service creates all topics.
Reviewers: Matthias J. Sax <mjsax@apache.org>, John Roesler <john@confluent.io>
Allow the Trogdor agent to execute external commands. The agent communicates with the external commands via stdin, stdout, and stderr.
Based on a patch by Xi Yang <xi@confluent.io>
Reviewers: David Arthur <mumrah@gmail.com>
* Enable heap dumps on OOM with -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=<file.bin> in the major services in system tests
* Collect the heap dump from the predefined location as part of the result logs for each service
* Change Connect service to delete the whole root directory instead of individual expected files
* Tested by running the full suite of system tests
Author: Konstantine Karantasis <konstantine@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#6158 from kkonstantine/KAFKA-7834
* Allow the Trogdor agent to be started in "exec mode", where it simply
runs a single task and exits after it is complete.
* For AgentClient and CoordinatorClient, allow the user to pass the path
to a file containing JSON, instead of specifying the JSON object in the
command-line text itself. This means that we can get rid of the bash
scripts whose only function was to load task specs into a bash string
and run a Trogdor command.
* Print dates and times in a human-readable way, rather than as numbers
of milliseconds.
* When listing tasks or workers, output human-readable tables of
information.
* Allow the user to filter on task ID name, task ID pattern, or task
state.
* Support a --json flag to provide raw JSON output if desired.
Reviewed-by: David Arthur <mumrah@gmail.com>, Stanislav Kozlovski <stanislav_kozlovski@outlook.com>
When using ./ducker-ak test on Jenkins, the script complains that there is no TTY. To fix this, we should skip passing -t to docker exec. We do not need a pseudo-TTY to run the tests. Similarly, we should skip passing -i, since we do not need to keep stdin open.
The down command should have a force option, specified as -f or --force.
Reviewed-by: Colin P. McCabe <cmccabe@apache.org>
[KIP-297](https://cwiki.apache.org/confluence/display/KAFKA/KIP-297%3A+Externalizing+Secrets+for+Connect+Configurations#KIP-297:ExternalizingSecretsforConnectConfigurations-PublicInterfaces) introduced the `ConfigProvider` mechanism, which was primarily intended for externalizing secrets provided in connector configurations. However, when querying the Connect REST API for the configuration of a connector or its tasks, those secrets are still exposed. The changes here prevent the Connect REST API from ever exposing resolved configurations in order to address that. rhauch has given a more thorough writeup of the thinking behind this in [KAFKA-5117](https://issues.apache.org/jira/browse/KAFKA-5117)
Tested and verified manually. If these changes are approved unit tests can be added to prevent a regression.
Author: Chris Egerton <chrise@confluent.io>
Reviewers: Robert Yokota <rayokota@gmail.com>, Randall Hauch <rhauch@gmail.com, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#6129 from C0urante/hide-provided-connect-configs
+ Add a parameter to the ducktap-ak to control the OpenJDK base image.
+ Fix a few issues of using OpenJDK:11 as the base image.
*More detailed description of your change,
if necessary. The PR title and PR message become
the squashed commit message, so use a separate
comment to ping reviewers.*
*Summary of testing strategy (including rationale)
for the feature or bug fix. Unit and/or integration
tests are expected for any behaviour change and
system tests should be considered for larger changes.*
Author: Xi Yang <xi@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#6071 from yangxi/ducktape-jdk
When I originally refactored the streams_upgrade_test#upgrade_downgrade_brokers test I removed the wait call which would wait for up 24 minutes for the SmokeTestDriver class to publish and verify all of its records.
Since most of the tests run in two minutes or less I set the monitor_log timeout to three minutes. However, the SmokeTestDriver#verify method allows up to six minutes to consume all records before verifying the monitor_log timeout needs to be greater than 6 minutes. I've set the timeout to 8 minutes.
Additionally, the steps needed to update the streams_upgrade_test should be documented as there are several components that need to get updated. So I've documented those steps here on the test as a giant comment.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This commit creates an EndToEndTest base class which relies on the verifiable consumer. This will ultimately replace ProduceConsumeValidateTest which depends on the console consumer. The advantage is that the verifiable consumer exposes more information to use for validation. It also allows for a nicer shutdown pattern. Rather than relying on the console consumer idle timeout, which requires a minimum wait time, we can halt consumption after we have reached the last acked offsets. This should be more reliable and faster. The downside is that the verifiable consumer only works with the new consumer, so we cannot yet convert the upgrade tests. This commit converts only the replication tests and a flaky security test to use EndToEndTest.
The StreamsUpgradeTest::test_upgrade_downgrade_brokers used sleep calls in the test which led to flaky test performance and as a result, we placed an @ignore annotation on the test. This PR uses log events instead of the sleep calls hence we can now remove the @ignore setting.
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>, Matthias J. Sax <matthias@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Some Streams system tests have failed during the setup phase
due to the producer having retries disabled and getting some
transient error from the broker.
This patch adds a retries parameter to the VerifiableProducer
(default unchanged), and sets retries to 10 for Streams tests.
It also sets acks equal to the number of brokers for Streams tests.
Reviewers: Matthias J. Sax <matthias@confluent.io>, Bill Bejeck <bill@confluent.io>, Guozhang Wang <guozhang@confluent.io>
Previous PR #6043 reduced throughput for VerifiableProducer in base class, but the streams_standby_replica_test needs higher throughput for consumer to complete verification in 60 seconds
Update system test and kicked off branch builder with 25 repeats https://jenkins.confluent.io/job/system-test-kafka-branch-builder/2201/
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This PR addresses a few issues with this system test flakiness. This PR is a cherry-picked duplicate of #6041 but for the trunk branch, hence I won't repeat the inline comments here.
1. Need to grab the monitor before a given operation to observe logs for signal
2. Relied too much on a timely rebalance and only sent a handful of messages.
I've updated the test and ran it here https://jenkins.confluent.io/job/system-test-kafka-branch-builder/2143/ parameterized for 15 repeats all passed.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This is the error message we're after:
"You may not specify an explicit partition assignment when using multiple consumers in the same group."
We apparently changed it midway through #5810 and forgot to update the test.
Reviewers: Dhruvil Shah <dhruvil@confluent.io>, Ismael Juma <ismael@juma.me.uk>
This is a system test for doing a rolling upgrade of a topology with a named repartition topic.
1. An initial Kafka Streams application is started on 3 nodes. The topology has one operation forcing a repartition and the repartition topic is explicitly named.
2. Each node is started and processing of data is validated
3. Then one node is stopped (full stop is verified)
4. A property is set signaling the node to add operations to the topology before the repartition node which forces a renumbering of all operators (except repartition node)
5. Restart the node and confirm processing records
6. Repeat the steps for the other 2 nodes completing the rolling upgrade
I ran two runs of the system test with 25 repeats in each run for a total of 50 test runs.
All test runs passed
Reviewers: John Roesler <john@confluent.io>, Matthias J. Sax <matthias@confluent.io>
This change adds some basic system tests for delegation token based authentication:
- basic delegation token creation
- producing with a delegation token
- consuming with a delegation token
- expiring a delegation token
- producing with an expired delegation token
New files:
- delegation_tokens.py: a wrapper around kafka-delegation-tokens.sh - executed in container where a secure Broker is running (taking advantage of automatic cleanup)
- delegation_tokens_test.py: basic test to validate the lifecycle of a delegation token
Changes were made in the following file to extend their functionality:
- config_property was updated to be able to configure Kafka brokers with delegation token related settings
- jaas.conf template because a broker needs to support multiple login modules when delegation tokens are used
- consule-consumer and verifiable_producer to override KAFKA_OPTS (to specify custom jaas.conf) and the client properties (to authenticate with delegation token).
Author: Attila Sasvari <asasvari@apache.org>
Reviewers: Reviewers: Viktor Somogyi <viktorsomogyi@gmail.com>, Andras Katona <41361962+akatona84@users.noreply.github.com>, Manikumar Reddy <manikumar.reddy@gmail.com>
Closes#5660 from asasvari/KAFKA-4544
This is a new system test testing for optimizing an existing topology. This test takes the following steps
1. Start a Kafka Streams application that uses a selectKey then performs 3 groupByKey() operations and 1 join creating four repartition topics
2. Verify all instances start and process data
3. Stop all instances and verify stopped
4. For each stopped instance update the config for TOPOLOGY_OPTIMIZATION to all then restart the instance and verify the instance has started successfully also verifying Kafka Streams reduced the number of repartition topics from 4 to 1
5. Verify that each instance is processing data from the aggregation, reduce, and join operation
Stop all instances and verify the shut down is complete.
6. For testing I ran two passes of the system test with 25 repeats for a total of 50 test runs.
All test runs passed
Reviewers: Matthias J. Sax <matthias@confluent.io>, Bill Bejeck <bill@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
KAFKA-7597: Add configurable transaction support to ProduceBenchWorker. In order to get support for serializing Optional<> types to JSON, add a new library: jackson-datatype-jdk8. Once Jackson 3 comes out, this library will not be needed.
Reviewers: Colin McCabe <cmccabe@apache.org>, Ismael Juma <ismael@juma.me.uk>
The Kafka Streams system tests fail with some regularity due to a timeout starting the broker.
The initial start is quite quick, but many of our tests involve stopping and restarting nodes with data already loaded, and also while processing is ongoing.
Under these conditions, it seems to be normal for the broker to take about 25 seconds to start, which makes the 30 second timeout pretty close for comfort.
I have seen many test failures in which the broker successfully started within a couple of seconds after the tests timed out and already initiated the failure/shut-down sequence.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
The function `setup_producer_and_consumer` is unused in the system test
framework, which incorrectly suggests subclasses should implement
it. It is not required or even referenced by the framework, so
the requirement should be removed.
Reviewers: Viktor Somogyi <viktorsomogyi@gmail.com>, Jason Gustafson <jason@confluent.io>
Add threads with separate consumers to ConsumeBenchWorker. Update the Trogdor test scripts and documentation with the new functionality.
Reviewers: Colin McCabe <cmccabe@apache.org>
Currently, the startup timeout is hardcoded to be 60 seconds in Connect's test service. Modifying it to be passable via init. This can safely be backported as well.
Reviewers: Randall Hauch <rhauch@gmail.com>, Jason Gustafson <jason@confluent.io>
This fixes the Connect standalone system tests. See branch builder: https://jenkins.confluent.io/job/system-test-kafka-branch-builder/2021/
This should be backported to the 2.0 branch, since that's when the tests were first
modified to use the external property file.
Reviewers: Magesh Nandakumar <magesh.n.kumar@gmail.com>, Ismael Juma <ismael@juma.me.uk>
Corrects an error in the system tests:
```
07:55:45 [ERROR:2018-10-23 07:55:45,738]: Failed to import kafkatest.tests.connect.connect_test, which may indicate a broken test that cannot be loaded: NameError: name 'EXTERNAL_CONFIGS_FILE' is not defined
```
The constant is defined in the [services/connect.py](https://github.com/apache/kafka/blob/trunk/tests/kafkatest/services/connect.py#L43) file in the `ConnectServiceBase` class, but the problem is in the [tests/connect/connect_test.py](https://github.com/apache/kafka/blob/trunk/tests/kafkatest/tests/connect/connect_test.py#L50) `ConnectStandaloneFileTest`, which does *not* extend the `ConnectServiceBase class`. Suggestions welcome to be able to reuse that variable without duplicating the literal (as in this PR).
System test run with this PR: https://jenkins.confluent.io/job/system-test-kafka-branch-builder/2004/
If approved, this should be merged as far back as the `2.0` branch.
Author: Randall Hauch <rhauch@gmail.com>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#5832 from rhauch/fix-connect-externals-tests
1. In test_upgrade_downgrade_brokers, allow duplicates to happen.
2. In test_version_probing_upgrade, grep the generation numbers from brokers at the end, and check if they can ever be synchronized.
Reviewers: John Roesler <john@confluent.io>, Matthias J. Sax <matthias@confluent.io>, Bill Bejeck <bill@confluent.io>
PR #2267 Introduced support for Zstandard compression. The relevant test expects values for `num_nodes` and `num_producers` based on the (now-incremented) count of compression types.
Passed the affected, previously-failing test:
`ducker-ak test tests/kafkatest/tests/client/compression_test.py`
Reviewers: Jason Gustafson <jason@confluent.io>
Pricing for m3.xlarge: On-Demand is at $0.266. Reserved is at about $0.16 (40% discount). And Spot is at $0.0627 (76% discount relative to On-Demand, or 60% discount relative to Reserved). Insignificant fluctuation in the past 3 months.
Ran on branch builder and works as expected -- each worker is created using spot instances (https://jenkins.confluent.io/job/system-test-kafka-branch-builder/1982/console)
This can be safely backported to 0.10.2 (tested using https://jenkins.confluent.io/job/system-test-kafka-branch-builder/1983/)
Author: Max Zheng <maxzheng.os@gmail.com>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#5707 from maxzheng/minor-switch@trunk
In some tests, the check monitoring the JMX tool log output doesn’t quite wait long enough before failing. Increasing the timeout from 10 to 20 seconds.
Removed ignore annotations from the upgrade tests. This PR includes the following changes for updating the upgrade tests:
* Uploaded new versions 0.10.2.2, 0.11.0.3, 1.0.2, 1.1.1, and 2.0.0 (in the associated scala versions) to kafka-packages
* Update versions in version.py, Dockerfile, base.sh
* Added new versions to StreamsUpgradeTest.test_upgrade_downgrade_brokers including version 2.0.0
* Added new versions StreamsUpgradeTest.test_simple_upgrade_downgrade test excluding version 2.0.0
* Version 2.0.0 is excluded from the streams upgrade/downgrade test as StreamsConfig needs an update for the new version, requiring a KIP. Once the community votes the KIP in, a minor follow-up PR can be pushed to add the 2.0.0 version to the upgrade test.
* Fixed minor bug in kafka-run-class.sh for classpath in upgrade/downgrade tests across versions.
* Follow on PRs for 0.10.2x, 0.11.0x, 1.0.x, 1.1.x, and 2.0.x will be pushed soon with the same updates required for the specific version.
Reviewers: Eno Thereska <eno.thereska@gmail.com>, John Roesler <vvcephei@users.noreply.github.com>, Guozhang Wang <wangguoz@gmail.com>, Matthias J. Sax <matthias@confluent.io>
Currently, the only way in system tests to add a new variable to the `jaas.conf` template file is to directly edit the path the config is constructed by adding new keyword arguments.
This wasn't necessarily a big problem, since you'd only need edit the `security_config.py` file as JAAS settings should come from the security settings.
Now, with the addition of [KIP-342](https://cwiki.apache.org/confluence/display/KAFKA/KIP-342%3A+Add+support+for+Custom+SASL+extensions+in+OAuthBearer+authentication), the OAuthBearer JAAS config supports arbitrary values in the form of SASL extensions. This patch exposes a more convenient API to overrides these values in system tests.
Reviewers: Jason Gustafson <jason@confluent.io>
Fix system tests from earlier #5445 by moving to the `ConnectSystemBase` class the creation & cleanup of a file that can be used as externalized secrets in connector configs.
Reviewers: Arjun Satish <arjun@confluent.io>, Robert Yokota <rayokota@gmail.com>, Konstantine Karantasis <konstantine@confluent.io>, Jason Gustafson <jason@confluent.io>
This is a fix to #5226 to account for config properties that have an
equal char in the value. Otherwise if there is one
equal char in the value the following error occurs:
dictionary update sequence element #XX has length 3; 2 is required
Reviewers: Colin Patrick McCabe <colin@cmccabe.xyz>, Ismael Juma <ismael@juma.me.uk>
* Updated TestLogCleaning tool to use Java consumer and rename as LogCompactionTester.
* Enabled the log cleaner in every system test.
* Removed configs from "kafka.properties" with default values and `socket.receive.buffer.bytes`
as the override did not seem necessary.
* Updated `kafka.py` logic to handle duplicates between `kafka.properties` and `server_prop_overrides`.
* Updated Gradle build so that classes from `kafka-clients` test jar can be used in
system tests.
Reviewers: Colin Patrick McCabe <colin@cmccabe.xyz>, Ismael Juma <ismael@juma.me.uk>
Added a system test which creates a file sink with json converter and attempts to feed it bad records. The bad records should land in the DLQ if it is enabled, and the task should be killed or bad records skipped based on test parameters.
Signed-off-by: Arjun Satish <arjunconfluent.io>
*More detailed description of your change,
if necessary. The PR title and PR message become
the squashed commit message, so use a separate
comment to ping reviewers.*
*Summary of testing strategy (including rationale)
for the feature or bug fix. Unit and/or integration
tests are expected for any behaviour change and
system tests should be considered for larger changes.*
Author: Arjun Satish <arjun@confluent.io>
Reviewers: Konstantine Karantasis <konstantine@confluent.io>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#5456 from wicknicks/error-handling-sys-test
If a property requires validation, it should be pretransformed if it is a variable reference, in order to have a value that will properly pass the validation.
Author: Robert Yokota <rayokota@gmail.com>
Reviewers: Randall Hauch <rhauch@gmail.com>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#5445 from rayokota/KAFKA-7225-pretransform-validated-props
The original way of stopping the minikdc process sometimes misfires because the process arg string is very long, and `ps` is not able to find the correct process. Using the `kill_java_processes` method is more reliable for finding and killing java processes.
In system tests, it is useful to have the thread dumps if a broker cannot be stopped using SIGTERM.
Reviewers: Xavier Léauté <xl+github@xvrl.net>, Ismael Juma <ismael@juma.me.uk>, Jason Gustafson <jason@confluent.io>
1. At the beginning of assign, we first check that all the non-repartition source topics are included in the metadata. If not, we log an error at the leader and set an error in the Assignment userData bytes, indicating that leader cannot complete assignment and the error code would indicate the root cause of it.
2. Upon receiving the assignment, if the error is not NONE the streams will shutdown itself with a log entry re-stating the root cause interpreted from the error code.
Author: tedyu <yuzhihong@gmail.com>
Reviewers: Matthias J. Sax <mjsax@apache.org>, Guozhang Wang <wangguoz@gmail.com>
Closes#5322 from tedyu/trunk
#5253 broke standby restoration for windowed stores.
Reviewers: Bill Bejeck <bill@confluent.io>, Matthias J. Sax <matthias@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Added two additional test cases to quota_test.py, which run between brokers and clients with different throttling behaviors. More specifically,
1. clients with new throttling behavior (i.e., post-KIP-219) and brokers with old throttling behavior (i.e., pre-KIP-219)
2. clients with old throttling behavior and brokers with new throttling behavior
Author: Jon Lee <jonlee@linkedin.com>
Author: Dong Lin <lindong28@gmail.com>
Reviewers: Dong Lin <lindong28@gmail.com>
Closes#5294 from jonlee2/kafka-6944
- Removed Scala consumers (`SimpleConsumer` and `ZooKeeperConsumerConnector`)
and their tests.
- Removed Scala request/response/message classes.
- Removed any mention of new consumer or new producer in the code
with the exception of MirrorMaker where the new.consumer option was
never deprecated so we have to keep it for now. The non-code
documentation has not been updated either, that will be done
separately.
- Removed a number of tools that only made sense in the context
of the Scala consumers (see upgrade notes).
- Updated some tools that worked with both Scala and Java consumers
so that they only support the latter (see upgrade notes).
- Removed `BaseConsumer` and related classes apart from `BaseRecord`
which is used in `MirrorMakerMessageHandler`. The latter is a pluggable
interface so effectively public API.
- Removed `ZkUtils` methods that were only used by the old consumers.
- Removed `ZkUtils.registerBroker` and `ZKCheckedEphemeral` since
the broker now uses the methods in `KafkaZkClient` and no-one else
should be using that method.
- Updated system tests so that they don't use the Scala consumers except
for multi-version tests.
- Updated LogDirFailureTest so that the consumer offsets topic would
continue to be available after all the failures. This was necessary for it
to work with the Java consumer.
- Some multi-version system tests had not been updated to include
recently released Kafka versions, fixed it.
- Updated findBugs and checkstyle configs not to refer to deleted
classes and packages.
Reviewers: Dong Lin <lindong28@gmail.com>, Manikumar Reddy <manikumar.reddy@gmail.com>
This could be backported to older branches to reduce the extra log warning messages there, too.
Running Connect system tests in this branch builder job: https://jenkins.confluent.io/job/system-test-kafka-branch-builder/1773/
Author: Randall Hauch <rhauch@gmail.com>
Reviewers: Konstantine Karantasis <konstantine@confluent.io>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#5151 from rhauch/kafka-7009
In BrokerCompatibilityTest.java, when older versioned broker is used (0.10.1, 0.10.2), LIST_OFFSET is not supported as well. Hence in the verification phase, there is a possibility that consumer hit the UnsupportedVersionException earlier than Streams actually hits it:
org.apache.kafka.common.errors.UnsupportedVersionException: The broker does not support LIST_OFFSETS with version in range [2,3]. The supported range is [0,1].
While the test is waiting for
org.apache.kafka.common.errors.UnsupportedVersionException: Cannot create a v0 FindCoordinator request because we require features supported only in 1 or later.
Both are valid errors to expect (the former is from consumer while the latter is from producer of the streams app).
Reviewers: John Roesler <john@confluent.io>, Matthias J. Sax <matthias@confluent.io>, Bill Bejeck <bill@confluent.io>
Reviewers: Viktor Somogyi <viktorsomogyi@gmail.com>, Vahid Hashemian <vahidhashemian@us.ibm.com>, Manikumar Reddy <manikumar.reddy@gmail.com>, Ismael Juma <ismael@juma.me.uk>
Implementation of [KIP-174](https://cwiki.apache.org/confluence/display/KAFKA/KIP-174+-+Deprecate+and+remove+internal+converter+configs+in+WorkerConfig)
Configuration properties 'internal.key.converter' and 'internal.value.converter'
are deprecated, and default to org.apache.kafka.connect.json.JsonConverter.
Warnings are logged if values are specified for either, or if properties that
appear to configure instances of internal converters (i.e., ones prefixed with
either 'internal.key.converter.' or 'internal.value.converter.') are given.
The property 'schemas.enable' is also defaulted to false for internal
JsonConverter instances (both for keys and values) if it isn't specified.
Documentation and code have also been updated with deprecation notices and
annotations, respectively.
Unit tests have been updated in `PluginsTest` to account for the new defaults for `schemas.enable` for internal key/value converters, and to ensure that (for the time being), internal key/value converters are still configurable despite being deprecated.
Author: Chris Egerton <chrise@confluent.io>
Author: Ewen Cheslack-Postava <me@ewencp.org>
Reviewers: Randall Hauch <rhauch@gmail.com>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#4693 from C0urante/kafka-5540
* Removed Scala producers, request classes, kafka.tools.ProducerPerformance, encoders,
tests.
* Updated ConsoleProducer to remove Scala producer support (removed `BaseProducer`
and several options that are not used by the Java producer).
* Updated a few Scala consumer tests to use the new producer (including a minor
refactor of `produceMessages` methods in `TestUtils`).
* Updated `ClientUtils.fetchTopicMetadata` to use `SimpleConsumer` instead of
`SyncProducer`.
* Removed `TestKafkaAppender` as it looks useless and it defined an `Encoder`.
* Minor import clean-ups
No new tests added since behaviour should remain the same after these changes.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Manikumar Reddy O <manikumar.reddy@gmail.com>, Dong Lin <lindong28@gmail.com>
Closes#5045 from ijuma/kafka-6921-remove-old-producer
test_broker_type_bounce_at_start tries to validate that when the controller is down, the streams client will always fail trying to create the topic; with the current behavior of admin client it is actually not always true: the actual behavior depends on the admin client internals as well as when the controller becomes unavailable during the leader assign partitions phase. I'd suggest at least ignore this test for now until the admin client has more stable (personally I'd even suggest removing this test as its coverage benefits is smaller than its introduced issues to me).
Also adding a few more log4j entries as a result of investigating this issue.
Reviewers: Matthias J. Sax <matthias@confluent.io>
Due to #4644 the consumer connector logs will be much more clean with fewer "broker may not be available" entries. We need to reduce the required frequency from 100 to a smaller number.
I've thought about reducing to just 1, but it may still be transient (i.e. even if broker is starting up you may see a few entries) so I reduced it to 10.
Reviewers: Bill Bejeck <bill@confluent.io>, Matthias J. Sax <matthias@confluent.io>
This PR does the following:
* Remove the StreamsRepeatingIntegerKeyProducerService and the associated Java class
* Add a parameter to VerifiableProducer.java to enable sending keys when specified
* Update the corresponding Python file verifiable_producer.py to support the new parameter.
Reviewers: Matthias J Sax <matthias@confluentio>, Guozhang Wang <wangguoz@gmail.com>
If not pinned, the following error will happen:
Traceback (most recent call last):
File "/usr/bin/pip", line 9, in <module>
from pip import main
ImportError: cannot import name main
Reviewers: Guozhang Wang <wangguoz@gmail.com>
This PR fixes some regressions introduced into streams system tests and sets the upgrade tests to ignore until PR #4636 is merged as it has the fixes for the upgrade tests.
Reviewers: Guozhang Wang <wangguoz@gmail.com>
There are a couple minor additions in this PR:
1. Add a new test for window store, to range query upon receiving each record.
2. In the non-windowed state store case, add a get call before the put call.
3. Enable caching by default to be consistent with other Join / Aggregate cases, where caching is enabled by default.
Reviewers: Bill Bejeck <bill@confluent.io>, Matthias J. Sax <matthias@confluent.io>
- adds Streams upgrade tests for 1.1 release
- introduces metadata version 3
Reviewers: John Roesler <john@confluent.io>, Guozhang Wang <guozhang@confluent.io>
SimpleBenchmark:
1.a Do not rely on manual num.records / bytes collection on atomic integers.
1.b Rely on config files for num.threads, bootstrap.servers, etc.
1.c Add parameters for key skewness and value size.
1.d Refactor the tests for loading phase, adding tumbling-windowed count.
1.e For consumer / consumeproduce, collect metrics on consumer instead.
1.f Force stop the test after 3 minutes, this is based on empirical numbers of 10M records.
Other tests: use config for kafka bootstrap servers.
streams_simple_benchmark.py: only use scale 1 for system test, remove yahoo from benchmark tests.
Note that the JMX based metrics is more accurate than the manually collected metrics.
Reviewers: John Roesler <john@confluent.io>, Bill Bejeck <bill@confluent.io>, Matthias J. Sax <matthias@confluent.io>
The standby-task test failed due to standby task distribution not be exactly equal. I think this will be the case from time to time, so I've updated test to make sure the standby task assignment count is not zero.
Reviewers: Guozhang Wang <wangguoz@gmail.com>, John Roesler <john@confluent.io>, Matthias J. Sax <matthias@confluent.io>
1. Use JmxMixin for SimpleBenchmark (will remove the self reporting in #4744), only when loading phase is false (i.e. we are in fact starting the streams app).
2. Reported the full jmx reported metrics in log files, and in the returned data only return the max values: this is because we want to skip the warming up and cooling down periods that will have lower rate numbers, while max represents the actual rate at full speed.
3. Incorporates two other improves to JMXTool: #1241 and #2950
Reviewers: John Roesler <john@confluent.io>, Matthias J. Sax <matthias@confluent.io>, Rohan Desai <desai.p.rohan@gmail.com>
This is a general change and is re-requisite to allow streams benchmark test with different streams tests. For the streams benchmark itself I will have a separate PR for switching configs. Details:
1. Create a "streams.properties" file under PERSISTENT_ROOT before all the streams test. For now it will only contain a single config of state.dir pointing to PERSISTENT_ROOT.
2. For all the system test related code, replace the main function parameter of state.dir with propsFilename, then inside the function load the props from the file and apply overrides if necessary.
3. Minor fixes.
Matthias J. Sax <matthias@confluent.io>, Bill Bejeck <bill@confluent.io>
Ensures Kafka chroot exists in ZK when starting KafkaService so commands that use ZK and are executed before the first Kafka broker starts do not fail due to the missing chroot.
Also uses chroot with one test that also has security parameterizations so Kafka's test suite exercises these combinations. Previously no tests were exercising chroots.
Changes were validated using sanity_checks which include the chroot-ed test as well as some non-chroot-ed tests.
Some changes required to get the Streams system tests working via Docker
To test:
TC_PATHS="tests/kafkatest/tests/streams" bash tests/docker/run_tests.sh
That command will take about 3.5 hours, and should pass. Note there are a couple of ignored tests.
Reviewers: Guozhang Wang <wangguoz@gmail.com>, Bill Bejeck <bill@confluent.io>
Author: Bill Bejeck <bill@confluent.io>
Reviewers: Damian Guy <damian@confluent.io>, Guozhang Wang <guozhang@confluent.io>, Matthias J. Sax <matthias@confluent.io>
ensure that brokers are registered at ZK before start() returns
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>, Damian Guy <damian@confluent.io>, Guozhang Wang <guozhang@confluent.io>
**NOTE: This should be backported to the `1.1` branch, and is currently a blocker for 1.1.**
The `connect_test.py::ConnectStandaloneFileTest.test_file_source_and_sink` system test is failing with the SASL configuration without a sufficient explanation. During the test, the Connect worker fails to start, but the Connect log contains no useful information. There are actual several things compounding to cause the failure and make it difficult to understand the problem.
First, the `tests/kafkatest/tests/connect/templates/connect_standalone.properties` is only adding in the broker's security configuration with the `producer.` and `consumer.` prefixes, but is not adding them with no prefix. The worker uses the AdminClient to connect to the broker to get the Kafka cluster ID and to manage the three internal topics, and the AdminClient is configured via top-level properties. Because the SASL test requires the clients all connect using SASL, the lack of broker security configs means the AdminClient was attempting and failing to connect to the broker. This is corrected by adding the broker's security configuration to the Connect worker configuration file at the top-level. (This was already being done in the `connect_distributed.properties` file.)
Second, the default `request.timeout.ms` for the AdminClient (and the other clients) is 120 seconds, so the AdminClient was retrying for 120 seconds before it would give up and thrown an error. However, the test was only waiting for 60 seconds before determining that the service failed to start. This can be corrected by setting `request.timeout.ms=10000` in the Connect distributed and standalone worker configurations.
Third, the Connect workers were recently changed to lookup the Kafka cluster ID before it started the herder. This is unlike the older uses of the AdminClient to find and manage the internal topics, where failure to connect was not necessarily logged correctly but nevertheless still skipped over, relying upon broker auto-topic creation to create the internal topics. (This may be why the test did not fail prior to the recent change to always require a successful AdminClient connection.) Although the worker never got this far in its startup process, the fact that we missed such an error since the prior releases means that failure to connect with the AdminClient was not being properly reported.
The `ConnectStandaloneFileTest.test_file_source_and_sink` system tests were run locally prior to this fix, and they failed as with the nightlies. Once these fixes were made, the locally run system tests passed.
Author: Randall Hauch <rhauch@gmail.com>
Reviewers: Konstantine Karantasis <konstantine@confluent.io>, Ewen Cheslack-Postava <me@ewencp.org>
Closes#4610 from rhauch/kafka-6577-trunk
fixes error message handling for test consumer client and KafkaStreams instance
updates expected error message
fixes race condition in system test code and avoids starting Streams processor twice
Author: Matthias J. Sax <matthias@confluent.io.>
Reviewer: Bill Bejeck <bill@confluent.io>, Guozhang Wang <guozhang@confluent.io>
Currently port forwarding is setup for HttpMetricsCollector when the Service's start_node method is called, but not canceled during stop. This hasn't presented a problem so far because we don't have tests that use this *and* restart the service. However, if a test/service does that, it will throw an exception since the port is already bound.
This just does the cleanup when stopping so a subsequent attempt to start again will succeed.
https://jenkins.confluent.io/job/system-test-kafka-branch-builder/1320 is a test run for a Test that uses ProducerPerformanceService, which in turn uses HttpMetricsCollector to validate the change.
Author: Ewen Cheslack-Postava <me@ewencp.org>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Apurva Mehta <apurva@confluent.io>
Closes#4604 from ewencp/cleanup-reverse-port-forward
Change the string in the test condition to the one that is logged
Author: Damian Guy <damian.guy@gmail.com>
Reviewers: Bill Bejeck <bill@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Closes#4599 from dguy/broker-compatibility
Another fall-through of `headers.converter` and `batch.size` properties. Here in `FileStreamSourceConnector` tests
Author: Konstantine Karantasis <konstantine@confluent.io>
Reviewers: Randall Hauch <rhauch@gmail.com>, Damian Guy <damian.guy@gmail.com>
Closes#4590 from kkonstantine/MINOR-Fix-file-source-task-config-in-system-tests
Avoid loosing log/stdout/stderr files on restart
Reenables tests
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Guozhang Wang <guozhang@confluent.io>, Bill Bejeck <bill@confluent.io>
* Rolling kill-restart Streams instances with brokers unavailable temporarily, and validate that the streams can still complete the restart process
Reviewers: Bill Bejeck <bill@confluent.io>, Matthias J. Sax <matthias@confluent.io>
System test for two cases:
* Starting a multi-node streams application with the broker down initially, broker starts and confirm rebalance completes and streams application still able to process records.
* Multi-node streams app running, broker goes down, stop stream instance(s) confirm after broker comes back remaining streams instance(s) still function.
Reviewers: Guozhang Wang <guozhang@confluent.io>, Matthias J. Sax <matthias@confluent.io>
Allow Connect Service in system tests to start asynchronously.
Specifically, allow for three startup conditions:
1. No condition - start async and return immediately.
2. Semi-async - start immediately after plugins have been discovered successfully.
3. Sync - start returns after the worker has completed startup. This is the current mode, but its condition is improved by checking that the port of Connect's REST interface is open, rather than that a log line has appeared in the logs.
An associated system test run has been started here:
https://jenkins.confluent.io/job/system-test-confluent-platform-branch-builder/586/
ewencp rhauch, I'd appreciate your review.
Author: Konstantine Karantasis <konstantine@confluent.io>
Reviewers: Randall Hauch <rhauch@gmail.com>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#4423 from kkonstantine/MINOR-Add-async-and-different-sync-startup-modes-in-ConnectService-test-class
We should start the process only within the `with` block, otherwise the bytes parameter would cause a race condition that result in false alarms of system test failures.
Author: Guozhang Wang <wangguoz@gmail.com>
Reviewers: Ewen Cheslack-Postava <me@ewencp.org>
Closes#4348 from guozhangwang/KMinor-fix-eos-test
* Implement process stop faults via SIGSTOP / SIGCONT
* Implement RoundTripWorkload, which both sends messages, and confirms that they are received at least once.
* Allow Trogdor tasks to block until other Trogdor tasks are complete.
* Add CreateTopicsWorker, which can be a building block for a lot of tests.
* Simplify how TaskSpec subclasses in ducktape serialize themselves to JSON.
* Implement some fault injection tests in round_trip_workload_test.py
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Rajini Sivaram <rajinisivaram@googlemail.com>
Closes#4323 from cmccabe/KAFKA-5849
Increase the number of messages produced to make the test more reliable. The test failed in a recent build and also fails intermittently when run locally. Since the producer uses acks=0 and the test stops as soon as a lag is observed, the change shouldn't have a big impact on the time taken to run when lag is observed sooner.
Author: Rajini Sivaram <rajinisivaram@googlemail.com>
Reviewers: Jason Gustafson <jason@confluent.io>
Closes#4312 from rajinisivaram/MINOR-replicaverification-test
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Bill Bejeck <bbejeck@gmail.com>, Guozhang Wang <wangguoz@gmail.com>
Closes#4242 from mjsax/kafka-4857-admit-client
Updated the System test `stream_broker_compatibility_test.py` to address system test failures as we have removed explicit broker version checking
- Ignore the `0.8.2.2` and `0.9.0.0` tests because the `NetworkClient` only logs `UnsupportedVersionException`s that occur and will continue to retry connecting. Once issue https://issues.apache.org/jira/browse/KAFKA-6297 is addressed, we may re-enable these tests.
- Updated existing tests expected error messages
- Updated Streams code in test for to make sure we fail fast for incompatible brokers
Author: Bill Bejeck <bill@confluent.io>
Reviewers: Matthias J. Sax <matthias@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Closes#4286 from bbejeck/MINOR_fix_broker_compatibility_tests
`topics.regex` was added in KAFKA-3073. This change fixes the test that invokes `/validate` to ensure that all the configdefs are returned as expected.
Author: Mikkin <mikkin@confluent.io>
Reviewers: Randall Hauch <rhauch@gmail.com>, Jason Gustafson <jason@confluent.io>, Ewen Cheslack-Postava <ewen@confluent.io>
Closes#4279 from mikkin/KAFKA-6284
Fix an omission where Kibosh was not getting installed on Vagrant
instances running in AWS.
Fix an issue where the Dockerfile was unable to download old Apache
Kafka releases. See the discussion on KAFKA-6233.
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Ismael Juma <ismael@juma.me.uk>
Closes#4240 from cmccabe/KAFKA-6247
For ducktape: add Kibosh to the testing Dockerfile.
Create files_unreadable_fault_spec.py.
For trogdor: create FilesUnreadableFaultSpec.java.
Add a unit test of using the Kibosh service.
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Rajini Sivaram <rajinisivaram@googlemail.com>
Closes#4195 from cmccabe/KAFKA-5811
Previously, Trogdor only handled "Faults." Now, Trogdor can handle
"Tasks" which may be either faults, or workloads to execute in the
background.
The Agent and Coordinator have been refactored from a
mutexes-and-condition-variables paradigm into a message passing
paradigm. No locks are necessary, because only one thread can access
the task state or worker state. This makes them a lot easier to reason
about.
The MockTime class can now handle mocking deferred message passing
(adding a message to an ExecutorService with a delay). I added a
MockTimeTest.
MiniTrogdorCluster now starts up Agent and Coordinator classes in
paralle in order to minimize junit test time.
RPC messages now inherit from a common Message.java class. This class
handles implementing serialization, equals, hashCode, etc.
Remove FaultSet, since it is no longer necessary.
Previously, if CoordinatorClient or AgentClient hit a networking
problem, they would throw an exception. They now retry several times
before giving up. Additionally, the REST RPCs to the Coordinator and
Agent have been changed to be idempotent. If a response is lost, and
the request is resent, no harm will be done.
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Rajini Sivaram <rajinisivaram@googlemail.com>, Ismael Juma <ismael@juma.me.uk>
Closes#4073 from cmccabe/KAFKA-6060
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Bill Bejeck <bill@confluent.io>, Jason Gustafson <jason@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Closes#4128 from mjsax/minor-cleanup
minor fix
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>, Alex Ayars <alex.ayars@confluent.io>
Closes#4084 from cmccabe/KAFKA-6070
With these changes, we are ensuring that the partitions being reassigned are from non-zero offsets. We also ensure that every message in the log has producerId and sequence number.
This means that it successfully reproduces https://issues.apache.org/jira/browse/KAFKA-6003.
Author: Apurva Mehta <apurva@confluent.io>
Reviewers: Jason Gustafson <jason@confluent.io>
Closes#4029 from apurvam/KAFKA-6016-add-idempotent-producer-to-reassign-partitions
- improve tests to get rid of calls to `sleep` in Python
- fixed some flaky test conditions
- improve debugging
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Damian Guy <damian.guy@gmail.com>, Bill Bejeck <bill@confluent.io>, Guozhang Wang <wangguoz@gmail.com>
Closes#3542 from mjsax/failing-eos-system-tests
To check ordering, we augment the existing transactions test to read and write from topics with one partition. Since we are writing monotonically increasing numbers, the topics should always be sorted, making it very easy to check for out of order messages.
Author: Apurva Mehta <apurva@confluent.io>
Reviewers: Jason Gustafson <jason@confluent.io>
Closes#3969 from apurvam/KAFKA-5888-system-test-which-check-ordering
The string representation of TopicPartition was changed to be
{topic}-{partitition} consistently in the following commit:
f6f56a645b
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Damian Guy <damian.guy@gmail.com>
Closes#3890 from ijuma/fix-replica-verification-test
Added some tips for running a single test file, test class and/or test method on the documentation landing page about tests
Author: Paolo Patierno <ppatierno@live.com>
Reviewers: Ismael Juma <ismael@juma.me.uk>
Closes#3577 from ppatierno/minor-tests-doc
Also throw an exception if a null keystore type is seen
in `SecurityStore`. This should never happen.
The default keystore type has changed in Java 9 (
http://openjdk.java.net/jeps/229), so we need to
be explicit to have consistent behaviour across
Java versions.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Rajini Sivaram <rajinisivaram@googlemail.com>
Closes#3808 from ijuma/set-jks-explicitly-in-system-tests
The previous timeout was 10 seconds, but system test failures have occurred when Zookeeper has started after about 11 seconds. Increasing the timeout to 30 seconds, since most of the time this extra time will not be required, and when it is it will prevent a failed system test.
In addition to merging to `trunk`, please backport to the `0.11.x` and `0.10.2.x` branches.
Author: Randall Hauch <rhauch@gmail.com>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#3774 from rhauch/MINOR-Increase-timeout-of-zookeeper-service-in-system-tests
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Apurva Mehta <apurva@confluent.io>, Ismael Juma <ismael@juma.me.uk>
Closes#3715 from cmccabe/kafka_service_print_node_hostname_on_failure
We are occasionally hitting some timeouts due to processing not finishing. So rather than failing the build for these reasons it would be better to reduce the runtime.
Author: Damian Guy <damian.guy@gmail.com>
Reviewers: Guozhang Wang <wangguoz@gmail.com>
Closes#3725 from dguy/fix-system-test
When a Connect distributed worker starts up talking with broker versions 0.10.1.0 and later, it will use the AdminClient to look for the internal topics and attempt to create them if they are missing. Although the AdminClient was added in 0.11.0.0, the AdminClient uses APIs to create topics that existed in 0.10.1.0 and later. This feature works as expected when Connect uses a broker version 0.10.1.0 or later.
However, when a Connect distributed worker starts up using a broker older than 0.10.1.0, the AdminClient is not able to find the required APIs and thus will throw an UnsupportedVersionException. Unfortunately, this exception is not caught and instead causes the Connect worker to fail even when the topics already exist.
This change handles the UnsupportedVersionException by logging a debug message and doing nothing. The existing producer logic will get information about the topics, which will cause the broker to create them if they don’t exist and broker auto-creation of topics is enabled. This is the same behavior that existed prior to 0.11.0.0, and so this change restores that behavior for brokers older than 0.10.1.0.
This change also adds a system test that verifies Connect works with a variety of brokers and is able to run source and sink connectors. The test verifies that Connect can read from the internal topics when the connectors are restarted.
Author: Randall Hauch <rhauch@gmail.com>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#3641 from rhauch/kafka-5704
ewencp would be great to cherry-pick this back into 0.11.x if possible
Author: Xavier Léauté <xavier@confluent.io>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#3645 from xvrl/system-test-cluster-id
Added handling of _DUCKTAPE_OPTIONS (mainly for enabling debugging)
Author: Paolo Patierno <ppatierno@live.com>
Reviewers: Ewen Cheslack-Postava <ewen@confluent.io>
Closes#3578 from ppatierno/kafka-5643
Support a --custom-ducktape flag which allows developers to install
their own versions of ducktape into Docker images. This is helpful for
ducktape development.
Author: Colin P. Mccabe <cmccabe@confluent.io>
Reviewers: Ewen Cheslack-Postava <me@ewencp.org>, Ismael Juma <ismael@juma.me.uk>
Closes#3539 from cmccabe/KAFKA-5602
The bump from 0.11.1.0-SNAPSHOT to 1.0.0-SNAPSHOT broke a couple
of system tests:
* TestVerifiableProducer.test_simple_run
* KafkaVersionTest.test_multi_version
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Damian Guy <damian.guy@gmail.com>
Closes#3587 from ijuma/fix-_kafka_jar_versions
Author: Dong Lin <lindong28@gmail.com>
Reviewers: Jiangjie Qin <becket.qin@gmail.com>, Jun Rao <junrao@gmail.com>, Ismael Juma <ismael@juma.me.uk>, Onur Karaman <okaraman@linkedin.com>
Closes#2929 from lindong28/KAFKA-4763
Waiting for the first line of output was added in KAFKA-2527 when JmxMixin was originally added as a heuristic to
determine when the process was ready. We've since determined this is not good enough given JmxTool's limitations
and now include a separate, more reliable check before starting JmxTool. This check is also dangerous since a
consumer that is started before data is available in the topic, it won't output anything to stdout and only logs
errors to a separate log file. This means we may have a long delay between starting the process and starting JMX
monitoring.
Since we have a more reliable check for liveness via JMX now (and in cases that need it, partition assignment
metrics via JMX), we should no longer need to wait for the first line of output.
Author: Ewen Cheslack-Postava <me@ewencp.org>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Apurva Mehta <apurva@confluent.io>
Closes#3447 from ewencp/dont-wait-first-line-console-consumer
Author: Matthias J. Sax <matthias@confluent.io>
Reviewers: Eno Thereska <eno.thereska@gmail.com>, Damian Guy <damian.guy@gmail.com>
Closes#3497 from mjsax/disable-flaky-system-tests
Update system tests to make use of the newly released 0.11 version.
Add on to https://github.com/apache/kafka/pull/3454
Author: Eno Thereska <eno.thereska@gmail.com>
Reviewers: Damian Guy <damian.guy@gmail.com>
Closes#3479 from enothereska/minor-compatibility-tests
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Eno Thereska <eno.thereska@gmail.com>, Ewen Cheslack-Postava <me@ewencp.org>
Closes#3454 from ijuma/test-upgrades-from-0.11.0.x
vinoth chandar
Brian Bushree
A. Sophie Blee-Goldman
Lucas Bradstreet
Bruno Cadonna
Manikumar Reddy
John Roesler
Randall Hauch
David Arthur
Jason Gustafson
Colin Patrick McCabe
Colin P. Mccabe
Boyang Chen
Bill Bejeck
Konstantine Karantasis
Bill Bejeck
Chris Egerton
Rajini Sivaram
Vikas Singh
vinoth chandar
Matthias J. Sax
Arjun Satish
Ismael Juma
Stanislav Vodetskyi
Stanislav Kozlovski
Alex Diachenko
Vahid Hashemian
Magesh Nandakumar
Guozhang Wang
Stanislav Kozlovski
Matthias J. Sax
Colin Patrick McCabe
Kan Li
Xi Yang
Attila Sasvari
Cyrus Vafadari
Gardner Vickers
Colin Hicks
Lee Dongjin
Max Zheng
Dong Lin
Anna Povzner
Robert Yokota
Arjun Satish
Jimmy Casey
Ted Yu
Jon Lee
Andy Coates
Paolo Patierno
Ewen Cheslack-Postava
Chris Egerton
Damian Guy
Mikkin
Adem Efe Gencer
Xavier Léauté
Magnus Edenhill
Apurva Mehta
Eno Thereska
Xavier Léauté