summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merge pull request #2018 from rabbitmq/rabbitmq-cli-302Michael Klishin2019-05-292-2/+26
|\ | | | | 'check' function for health check of virtual hosts
| * 'check' function for health check of virtual hostsDiana Corbacho2019-05-282-2/+26
| | | | | | | | [#163451547]
* | Merge pull request #2015 from Invizory/dead-letter-rejected-publishedDaniil Fedotov2019-05-298-34/+130
|\ \ | | | | | | Add an option to dead-letter rejected publishes
| * | Add remaining tests for `reject-publish-dlx`Arthur Khashaev2019-05-283-15/+43
| | | | | | | | | | | | | | | | | | | | | | | | Add tests for `reject-publish-dlx` overflow strategy for all corresponding tests for `reject-publish` strategy. The tests are grouped to avoid code duplication. See #1443
| * | Add overflow tests to `confirms_rejects_SUITE`Arthur Khashaev2019-05-281-13/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change splits the tests into groups for both the existing `reject-publish` overflow strategy and the new `reject-publish-dlx`. The corresponding queue names have been suffixed to make the tests in the groups independent and thus parallel. See #1443
| * | Add an option to dead-letter rejected publishesArthur Khashaev2019-05-274-6/+52
|/ / | | | | | | | | | | | | Add `reject-publish-dlx` overflow strategy, which is similar to `reject-publish` strategy, but also dead-letters rejected messages. Closes #1443
* | Improving confirms test stability and error reportingDaniil Fedotov2019-05-271-3/+10
| | | | | | | | | | | | Add process messages to the error message. Try harder to kill queues.
* | Improve error handling and increase timeout in confirm-reject testDaniil Fedotov2019-05-271-3/+4
| | | | | | | | | | This is to prevent false-positives on slow systems and detect genuine errors better.
* | Merge pull request #2014 from rabbitmq/quorum-statusMichael Klishin2019-05-231-9/+41
|\ \ | | | | | | Improve quorum_status command
| * | Improve quorum_status commandkjnilsson2019-05-231-9/+41
| | | | | | | | | | | | Make it compatible with the pretty table formatter.
* | | Update rabbitmq-components.mkJean-Sébastien Pédron2019-05-171-1/+1
| | |
* | | Update rabbitmq-components.mkMichael Klishin2019-05-171-1/+1
| | |
* | | Update rabbitmq-components.mkMichael Klishin2019-05-161-2/+2
| | |
* | | Merge pull request #2011 from ↵Michael Klishin2019-05-151-2/+2
|\ \ \ | | | | | | | | | | | | | | | | rabbitmq/fix-notify-limiter-after-consumer-timeout-change Adapt clause in notify_limiter to consumer timeout changes
| * | | Adapt clause in notify_limiter to consumer timeout changesArnaud Cogoluègnes2019-05-151-2/+2
|/ / /
* | | Update rabbitmq-components.mkMichael Klishin2019-05-141-1/+1
| | |
* | | Add a dependency on observer_cli, closes #2006Michael Klishin2019-05-131-1/+1
| | |
* | | Update rabbitmq-components.mkMichael Klishin2019-05-131-0/+1
| | |
* | | Merge pull request #1893 from rabbitmq/do_not_confirm_on_unreachable_queueMichael Klishin2019-05-134-244/+478
|\ \ \ | | | | | | | | Make sure that publishes to dead or unaccessible queues return nack
| * | | CosmeticsMichael Klishin2019-05-131-1/+1
| | | |
| * | | CosmeticsMichael Klishin2019-05-131-4/+4
| | | |
| * | | CosmeticsMichael Klishin2019-05-131-50/+38
| | | |
| * | | Change publisher confirms behaviour to reject messages if no queues confirmed.Daniil Fedotov2019-05-104-242/+488
|/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Channel is counting unacked messages in a remove-only data structure `dtree`. Each published message id is associated with a list of queues where it was routed to. On confirm of queue failures queues were removed from the list As soon as there are no queues in the list - the message can be confirmed. This meant that if all queues fail with "not abnormal" reasons - the message may be confirmed, but not enqueued. This change removes dtree data structure, replacing it with specific unconfirmed_messages data structure. It tracks queue pids similarly to dtree, but also has an API to record confirms and failures differently, keeping track of which queues received at least one confirm. If all pids fails or confirm, but not all queues received confirmation - it means not all queues enqueued the message and the message should be rejected This is different from the current behaviour, but corresponds to the docs and common sense. [#163952410]
* | | Merge pull request #2001 from rabbitmq/rabbitmq-server-2000v3.8.0-beta.4Michael Klishin2019-05-103-36/+80
|\ \ \ | |_|/ |/| | Move check for active readers to message store GC action function
| * | Compaction: more logging, a testMichael Klishin2019-05-102-2/+27
| | |
| * | DittoMichael Klishin2019-05-051-1/+1
| | |
| * | Log deferred file compaction at debug levelMichael Klishin2019-05-051-1/+1
| | |
| * | Move check for reader to action function for message store GC.Daniil Fedotov2019-05-032-34/+53
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | Message store GC postpones processing of file, which have readers. When performing an action, it asserts that there are no readers. Check for readers may race with readers update by a queue, crashing the message store. Make check and assert work with the same lookup to reduce failure rate. In case of races the queue process should handle exception instead. Addresses #2000 [#165755203]
* | Merge pull request #2004 from rabbitmq/send-drained-type-fixMichael Klishin2019-05-082-3/+3
|\ \ | |/ |/| Fix type issues with send_drained message
| * Fix type issues with send_drained messagekjnilsson2019-05-072-3/+3
|/ | | | | | For quorum queues. [#165796741]
* feature_flags_SUITE: Bump testcase timeout to 10 minutesJean-Sébastien Pédron2019-05-031-1/+1
|
* rabbit_feature_flags: Export initialize_registry/1 for the testsuiteJean-Sébastien Pédron2019-05-031-2/+2
| | | | | | | | Hiding the export behind the `TEST` macro should work on the paper. Unfortunately, when we compile a test plugin as part of the `feature_flags_SUITE` testsuite, `rabbit` is recompiled without the `TEST` macro. This causes the testsuite to fail with an `undef` exception because initialize_registry/1 is not exported anymore.
* Merge pull request #1998 from rabbitmq/consumer_timeouts_take_2Michael Klishin2019-05-033-61/+23
|\ | | | | Always close channel on consumer timeout
| * Log consumer timeout as a warningMichael Klishin2019-05-031-4/+4
| |
| * Drive by change: squash a compiler warningMichael Klishin2019-05-031-1/+0
| |
| * When consumer hits an ack timeout, log timeout value usedMichael Klishin2019-05-031-3/+3
| |
| * formattingkjnilsson2019-05-011-1/+1
| |
| * Always close channel on consumer timeoutkjnilsson2019-05-012-56/+19
| | | | | | | | | | | | In order to keep things simple and consistent this changes consumer timeouts to always close the channel instead of cancelling and returning.
* | rabbit_queue_location_validator: Fix pattern in validate_policy/1Jean-Sébastien Pédron2019-05-031-1/+1
| | | | | | | | | | | | | | | | validate_strategy/1 error value is `{error, FormatString, Args}`, not the usual `{error, Reason}`. This incorrect pattern matching was found thanks to the `set_policy_command_test.exs` testsuite in rabbitmq-cli.
* | Merge pull request #1999 from rabbitmq/queue-master-location-validationMichael Klishin2019-05-032-1/+18
|\ \ | | | | | | Policy validation must return 'ok' or an error
| * | Wording, explainMichael Klishin2019-05-031-2/+2
| | |
| * | Test invalid policy returns an errorDiana Corbacho2019-05-021-0/+14
| | | | | | | | | | | | [#165689664]
| * | Policy validation must return 'ok' or an errorDiana Corbacho2019-05-021-1/+4
| |/ | | | | | | | | | | Any other value will break the validation cycle and leave the rest of the arguments without validate [#165689664]
* | feature_flags_SUITE: Do not depend on the `quorum_queue` feature flagJean-Sébastien Pédron2019-05-021-73/+92
|/ | | | | This will ease the backport to 3.7.x where that feature flag does not exist.
* Update rabbitmq-components.mkGerhard Lazu2019-05-011-1/+1
|
* Update rabbitmq-components.mkLuke Bakken2019-04-301-1/+1
|
* rabbit_ff_extra: Use stdout_formatter to display `info()` tableJean-Sébastien Pédron2019-04-303-81/+98
| | | | [#164931055]
* Merge pull request #1995 from rabbitmq/sac-qq-fixArnaud Cogoluègnes2019-04-302-34/+74
|\ | | | | QQ SAC: process all consumers on noconnection
| * QQ SAC: process all consumers on noconnectionkjnilsson2019-04-302-34/+74
|/ | | | | | | | There can be multiple single active consumers in the consumers map if all but one are cancelled. Take this into account when processing noconnections. [#165438843]
* Merge pull request #1994 from rabbitmq/rabbitmq-server-1980Luke Bakken2019-04-292-3/+109
|\ | | | | Reset overflow queue property to drop-head if policy is unset.
View Lorry Controller Status