aiquestion commented on PR #11451: URL: https://github.com/apache/kafka/pull/11451#issuecomment-1121156225
> Thanks for reply. Will try to submit a PR for it. Yes, with [KAFKA-12984](https://issues.apache.org/jira/browse/KAFKA-12984), [KAFKA-13406](https://issues.apache.org/jira/browse/KAFKA-13406) rebalance will not stuck. What we see is that: with many consumers, the group may rebalance for many rounds before stable ( Cooperative rebalance enabled) * consumer A joined group G, and joinGroup completed **( with ownedPartition P1/P2 )** * consumer B joined group G, the group G state change to preparingRebalance * consumer A send syncGroup, and got REBALANCE_IN_PROGRESS error * consumer A rejoin group, send joinGroup to group G * consumer B send joinGroup to group G * consumer A and B complete the joinGroup and syncGroup successfully --- added --- * consumer A assigned partition P3/P4 (since it's ownedParititon ignored because of generation ) * consumer A revoke P1/P2 and send joinGroup for another round of rebalance Another round of rebalance begin and some other consumer C will not be able to syncGroup in time, so the rebalance will goes for many rounds before stable, and there will be dup consuming in the rebalance time. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
