Fix flake in rabbit_stream_queue_SUITE
This commit is contained in:
parent
c36b56b626
commit
1825f2be77
|
|
@ -170,7 +170,7 @@ all_tests_4() ->
|
|||
init_per_suite(Config0) ->
|
||||
rabbit_ct_helpers:log_environment(),
|
||||
Config = rabbit_ct_helpers:merge_app_env(
|
||||
Config0, {rabbit, [{stream_tick_interval, 1000},
|
||||
Config0, {rabbit, [{stream_tick_interval, 256},
|
||||
{log, [{file, [{level, debug}]}]}]}),
|
||||
rabbit_ct_helpers:run_setup_steps(Config).
|
||||
|
||||
|
|
@ -436,7 +436,11 @@ declare_queue(Config) ->
|
|||
declare(Config, Server, Q, [{<<"x-queue-type">>, longstr, <<"stream">>}])),
|
||||
|
||||
%% Test declare an existing queue
|
||||
?assertEqual({'queue.declare_ok', Q, 0, 0},
|
||||
%% there is a very brief race condition in the osiris counter updates that could
|
||||
%% cause the message count to be reported as 1 temporarily after a new stream
|
||||
%% creation. Hence to avoid flaking we don't match on the messages counter
|
||||
%% here
|
||||
?assertMatch({'queue.declare_ok', Q, _, 0},
|
||||
declare(Config, Server, Q, [{<<"x-queue-type">>, longstr, <<"stream">>}])),
|
||||
|
||||
?assertMatch([_], find_queue_info(Config, [])),
|
||||
|
|
|
|||
Loading…
Reference in New Issue