rkhachatryan commented on PR #20091: URL: https://github.com/apache/flink/pull/20091#issuecomment-1175451454
Regarding `TRIGGER_CHECKPOINT_FAILURE`, I **had** the following concerns - but after checking the code they turned out to be wrong: If counted as failure: - "not all tasks are running" could cause it - wrong: there is a separate `NOT_ALL_REQUIRED_TASKS_RUNNING` constant - timing issues or existing concurrent checkpoints can produce it - wrong: they are separate constants as well If not counted as failure (left as is): - if we check `mkdirs` return value then it will cause `TRIGGER_CHECKPOINT_FAILURE`, again breaking the counter - wrong: `mkdirs` is called during the `CheckpointCoordinator` initialization Still, I'm leaning towards leaving it as is because there are probably other edge cases when it happens and shouldn't be counted as a failure, e.g. shutdown. -- 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: issues-unsubscr...@flink.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
