gaoyunhaii commented on pull request #14820: URL: https://github.com/apache/flink/pull/14820#issuecomment-855904191
Hi @dawidwys I think this might be still happen in AC since 1. For the final version not all tasks might need to wait for the last aligned checkpoint since some tasks do not have operators using 2pc. 2. There might be still some cases even if `A` and `B` all need to wait for the last checkpoints. Suppose they are all waiting for checkpoint 4, with concurrent checkpoints we might start triggering checkpoint 5. However the network stalled after emitting barrier 4. Then we finally have checkpoint 4 completed, notify `A` and `B` to make them emit `EndOfPartition` and exit. Then if JM trigger checkpoint 6 now we would have the similar case. -- 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. For queries about this service, please contact Infrastructure at: [email protected]
