zaynt4606 commented on PR #3860: URL: https://github.com/apache/celeborn/pull/3860#issuecomment-6308003005
Thanks @SoulKyu for the fix. I pushed follow-up commit 80eae49cf to address the cleanup-order issue raised by @RexXiong and help move this PR forward. The follow-up preserves your workerGracefulShutdown guard and removes committedFileInfos entries after writer cleanup, so an in-flight close cannot reinsert metadata for an expired shuffle. I also expanded StorageManagerSuite to cover disabled/enabled graceful shutdown and a deterministic concurrent close/cleanup interleaving, and replaced the test's JIRA placeholder with CELEBORN-2495. The PR description now reflects the final behavior and validation. Validation: all 6 StorageManagerSuite tests passed locally on JDK 17. The concurrent cleanup regression fails against the previous PR head and passes with this fix. dev/reformat completed successfully for all 16 invocations. Please take another look when you have a chance. -- 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]
