[FLINK-22434] Dispatcher does not store suspended jobs in execution graph store

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.14.0, 1.13.1, 1.12.5
• Component/s: Runtime / Coordination
• Labels: pull-request-available
• Assignee: Fabian Paul
• Reporter: Fabian Paul
• Created: 23/Apr/21 15:35
• Updated: 09/Dec/21 11:28
• Resolved: 21/May/21 12:53

Description

Only globally terminated jobs are currently stored in the execution graph store after termination. In case the JobManager is shutdown and jobs are still running, these jobs will be suspended which is a non-globally terminated state.

The problem surfaces when a user tries to access information about the job during termination, leading to a job not found response. By storing all terminated jobs in the execution graph store this should be fixed.

Issue Links

• causes: FLINK-22820 Stopping Yarn session cluster will cause fatal error
• causes: FLINK-24232 Archiving of suspended jobs prevents breaks subsequent archive attempts
• links to: GitHub Pull Request #15799