zhuzhurk opened a new pull request #14405:
URL: https://github.com/apache/flink/pull/14405
## What is the purpose of the change
If user manually cancels a job when the job is failing(here failing means
the job encounters unrecoverable failure and is about to fail), the job will
hang in CANCELING state and cannot terminate. The cause is that
DefaultScheduler currently will always try to transition from `FAILING` to
`FAILED` to terminate the job. However, job canceling will change job status to
`CANCELING` so that the transition to `FAILED` will not success. This PR is to
fix this problem.
## Verifying this change
ExecutionGraphRestartTest is reworked to be based on the new scheduler to
verify it.
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): (yes / **no**)
- The public API, i.e., is any changed class annotated with
`@Public(Evolving)`: (yes / **no**)
- The serializers: (yes / **no** / don't know)
- The runtime per-record code paths (performance sensitive): (yes / **no**
/ don't know)
- Anything that affects deployment or recovery: JobManager (and its
components), Checkpointing, Kubernetes/Yarn/Mesos, ZooKeeper: (**yes** / no /
don't know)
- The S3 file system connector: (yes / **no** / don't know)
## Documentation
- Does this pull request introduce a new feature? (yes / **no**)
- If yes, how is the feature documented? (**not applicable** / docs /
JavaDocs / not documented)
----------------------------------------------------------------
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]
