lianetm commented on PR #15961: URL: https://github.com/apache/kafka/pull/15961#issuecomment-2149990136
Hey @appchemist, agree that as long as we continue polling the ConsumerNetworkThread we should get the error thrown (on the next poll after it was detected), but the inverted order we get with the managers approach (maybeThrowAnyException and then poll to detect the error to throw) would matter in cases when there's no such next poll right? I'm thinking about close for example, that we do one last poll (may detect metadata errors) and then we don't poll again (so it won't be thrown). The legacy consumer wouldn't behave like that I expect. Makes sense? If we agree that the managers approach seems a bit complicated and has an inverted order that brings a difference, why do we still prefer it over the simpler approach that is closer to the legacy logic? Do you see a downside to the simpler approach? (I may be missing it). Thanks! -- 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
