chia7712 commented on PR #16686: URL: https://github.com/apache/kafka/pull/16686#issuecomment-2423480587
> I think that we should also clarify how we handle interruption because this is actually the major difference between the old and the new implementation. In the old, we send the request before checking the interrupt flag whereas we implicitly check it before in the new implementation. This is actually why we saw members leaving properly in our tests with the old consumer even though close is still interrupted. I suppose that it is a gray area but let’s try to have a clear statement on it. As @hachikuji (https://github.com/apache/kafka/pull/16686#issuecomment-2411832933) and @kirktrue (https://github.com/apache/kafka/pull/16686#issuecomment-2423208911) mentioned in their comments, I don't think the old implementation guaranteed sending LEAVE_GROUP, and in my opinion, it's still a reasonable best-effort approach. > Agreed. That's why [KAFKA-17519](https://issues.apache.org/jira/browse/KAFKA-17519) was created; we need to define the intended behavior for all these different edge cases. If I recall correctly, @kirktrue raised a good question in KIP-1092 (https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=321719077) about the guarantee of sending LEAVE_GROUP. Maybe we could enhance the close option by adding a flag to ensure LEAVE_GROUP is sent, even if it means ignoring timeout and interruption? -- 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
