RexXiong commented on PR #3650: URL: https://github.com/apache/celeborn/pull/3650#issuecomment-4244439384
I think Copilot's suggestion makes sense here. When `previousFailureCount < 0` (reflection fails), checking `!hasRunningAttempt` before returning `true` would avoid premature stage reruns while still ensuring the failure is eventually reported. Could you elaborate on why "app will fail" if we check `!hasRunningAttempt`? If there's no running attempt, the method would still return `true` and trigger `FetchFailed`. The only difference is that when another attempt is running, we defer the failure report to give that attempt a chance to succeed. This seems consistent with the original logic where `!hasRunningAttempt` is one of the conditions to trigger `FetchFailed`. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
