[GitHub] [kafka] guozhangwang commented on pull request #11868: KAFKA-12648: fix flaky #shouldAddToEmptyInitialTopologyRemoveResetOffsetsThenAddSameNamedTopologyWithRepartitioning

2022-03-12 Thread GitBox
guozhangwang commented on pull request #11868: URL: https://github.com/apache/kafka/pull/11868#issuecomment-1066020270 > I think we need to document that these calls need to be made in the same order with the same topologies for each client. That is what KSQL does to make sure it works

[GitHub] [kafka] guozhangwang commented on pull request #11868: KAFKA-12648: fix flaky #shouldAddToEmptyInitialTopologyRemoveResetOffsetsThenAddSameNamedTopologyWithRepartitioning

2022-03-10 Thread GitBox
guozhangwang commented on pull request #11868: URL: https://github.com/apache/kafka/pull/11868#issuecomment-1064634922 @wcarlson5 @ableegoldman I'm wondering if we could still have live lock scenarios like this: say we have two topologies A and B, and two threads a and b each on a differen

[GitHub] [kafka] guozhangwang commented on pull request #11868: KAFKA-12648: fix flaky #shouldAddToEmptyInitialTopologyRemoveResetOffsetsThenAddSameNamedTopologyWithRepartitioning

2022-03-09 Thread GitBox
guozhangwang commented on pull request #11868: URL: https://github.com/apache/kafka/pull/11868#issuecomment-1063695247 > Anyways I'll push a fix for this, and then we should be good to go 👍 Thanks @wcarlson5 for the report! @ableegoldman please feel free to move on afterwards. --