[FLINK-31077] Trigger checkpoint failed but it were shown as COMPLETED by rest API

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.17.0, 1.15.3, 1.16.1
• Fix Version/s: 1.17.0, 1.16.2
• Component/s: Runtime / Checkpointing
• Labels: pull-request-available
• Assignee: Junrui Li
• Reporter: Junrui Li
• Created: 15/Feb/23 06:09
• Updated: 27/Feb/23 07:09
• Resolved: 27/Feb/23 07:09

Description

Currently, we can trigger a checkpoint and poll the status of the checkpoint until it is finished by rest according to FLINK-27101. However, even if the checkpoint status returned by rest is completed, it does not mean that the checkpoint is really completed. If an exception occurs after marking the pendingCheckpoint completed(here), the checkpoint is not written to the HA service and we can not failover from this checkpoint.

Issue Links

• is related to: FLINK-27101 Add REST API to manually trigger checkpoints
• links to: GitHub Pull Request #21943