sollhui opened a new pull request, #66105:
URL: https://github.com/apache/doris/pull/66105
### What problem does this PR solve?
Issue Number: None
Related PR: None
Problem Summary: Routine load task submission failures were either renewed
immediately or left running until timeout for resource pressure errors. The
job
therefore stayed `RUNNING` and users could miss a persistent failure.
This change pauses an active job with `CREATE_TASKS_ERR` for every task
submission failure, making the failure visible in the job state.
`CREATE_TASKS_ERR` remains eligible for the existing auto-resume policy, so
retryable failures are retried with the existing backoff instead of being
immediately requeued. Stale task callbacks and callbacks for jobs that are no
longer running remain ignored.
### Release note
Routine load jobs now enter `PAUSED` when a task cannot be submitted and rely
on auto resume for retryable recovery.
### Check List (For Author)
- Test
- [ ] Regression test
- [ ] Unit Test
- [ ] Manual test
- [x] No local test completed; local FE unit-test compilation was stopped
at the user's request. Unit coverage was updated for network,
`TOO_MANY_TASKS`, and `MEM_LIMIT_EXCEEDED` failures.
- Behavior changed:
- [x] Yes. Task submission failures pause the job instead of immediately
renewing the task or waiting for task timeout.
- Does this need documentation?
- [x] No.
### Check List (For Reviewer who merge this PR)
- [ ] Confirm the release note
- [ ] Confirm test cases
- [ ] Confirm document
- [ ] Add branch pick label
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
