Details
-
Improvement
-
Status: Closed
-
Critical
-
Resolution: Fixed
-
1.6.0, 1.7.0
-
The {{StandaloneJobClusterEntrypoint}} now starts all jobs with a fixed {{JobID}}. Thus, in order to run a cluster in HA mode, one needs to set a different cluster id {{high-availability.cluster-id}} for each job/cluster.
Description
The StandaloneJobClusterEntrypoint currently generates the JobGraph from the user code when being started. Due to the nature of how the JobGraph is generated, it will get a random JobID assigned. This is problematic in case of a failover because then, the JobMaster won't be able to detect the checkpoints. In order to solve this problem, we need to either fix the JobID assignment or make it configurable.
Attachments
Issue Links
- is related to
-
FLINK-10292 Generate JobGraph in StandaloneJobClusterEntrypoint only once
- Open
- links to