lianetm commented on PR #16686: URL: https://github.com/apache/kafka/pull/16686#issuecomment-2310866528
> the proposed behavior would be divergent from the existing consumer’s behavior uhm why? I could be missing something, but actually I see it as the proposed behaviour is to make both consumer similar on close+interrupt: send leave request and do not wait for broker responses. That's what the classic consumer achieves by throwing the interrupt right after polling the network client. With the current shape of the PR, we don't achieve the same with the new consumer, and actually, we introduce an unwanted behaviour imo: interrupted thread + close sits and wait for broker responses...response for the leave request, or any other in-flight request there may be. Is there an issue/concern with the alternative of calling close(0) if interrupted? How does the IT behave if we try the change? -- 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
