pnowojski commented on a change in pull request #14632: URL: https://github.com/apache/flink/pull/14632#discussion_r557325687
########## File path: docs/release-notes/flink-1.12.md ########## @@ -163,4 +163,10 @@ Beginning from Flink 1.12, jobs will be scheduled in the unit of pipelined regio The old scheduler can be enabled using the `jobmanager.scheduler.scheduling-strategy: legacy` setting. +### Known Issues +#### Unaligned checkpoint recovery may lead to corrupted data stream [FLINK-20654](https://issues.apache.org/jira/browse/FLINK-20654) + +Using UnalignedCheckpoints in Flink 1.12.0 combined with two/multiple inputs tasks or with union inputs for single input tasks can result in corrupted state. + +This can happen if a new checkpoint is triggered before recovery is fully completed. For state to be corrupted a task with two or more input gates must receive a checkpoint barrier exactly at the same time this tasks finishes recovering spilled in-flight data. In such case this new checkpoint can succeed, with corrupted/missing in-flight data, which will result in various deserialisation/corrupted data stream errors when someone attempts to recover from such corrupted checkpoint. Review comment: I would put it on top ---------------------------------------------------------------- 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]
