You are viewing a plain text version of this content. The canonical link for it is here.

Posted to jira@kafka.apache.org by "Colin McCabe (Jira)" <ji...@apache.org> on 2024/09/10 18:33:00 UTC

[jira] [Assigned] (KAFKA-17506) KRaftMigrationDriver initialization race condition

     [ https://issues.apache.org/jira/browse/KAFKA-17506?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Colin McCabe reassigned KAFKA-17506:
------------------------------------

    Affects Version/s: 3.7.1
                       3.8.0
                       3.9.0
             Assignee: David Arthur
             Priority: Blocker  (was: Major)

> KRaftMigrationDriver initialization race condition
> --------------------------------------------------
>
>                 Key: KAFKA-17506
>                 URL: https://issues.apache.org/jira/browse/KAFKA-17506
>             Project: Kafka
>          Issue Type: Bug
>    Affects Versions: 3.8.0, 3.7.1, 3.9.0
>            Reporter: David Arthur
>            Assignee: David Arthur
>            Priority: Blocker
>             Fix For: 3.9.0, 3.7.2, 3.8.1
>
>
> There is a race condition between KRaftMigrationDriver running its first poll() and being notified by Raft about a leader change. If {{onControllerChange}} is called before RecoverMigrationStateFromZKEvent is run, we will end up getting stuck in the INACTIVE state.
> This is likely a very narrow race condition that is not likely to be seen in practice. It was found when debugging a flaky integration test.
> If encountered, this can be worked around by restarting the stuck controller. Leadership will move to another controller which will allow the migration to be started.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)