gaoyunhaii commented on pull request #16905: URL: https://github.com/apache/flink/pull/16905#issuecomment-903942473
Very thanks Piotr @pnowojski for the review and explanation! Now I got the reason for asserting no exceptions~ > In that case I would propose to add a logic to differentiate those two cases. assertNoException for the first one, Environment#declineCheckpoint or StreamTask#declineCheckpoint in the latter one. Or if it makes sense AND the task is left in a valid state, always declineCheckpoint. However if an unexpected exception is thrown, failing hard might be a better way to go. I agree with we could distinguish the two cases. I found that originally we should also want to implement this logic in `assertTriggeringCheckpointExceptions`: the second case is identified by get RejectedExecutorException from the mailbox. But since the exception might be wrapped by CompletionException, thus I added the assertion back and slightly adjust its logic~ -- 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
