richardmilles opened a new pull request, #18469:
URL: https://github.com/apache/dolphinscheduler/pull/18469

   ## Was this PR generated or assisted by AI?
   
   YES. AI assistance was used for issue triage, locating the SeaTunnel cancel 
path, drafting the change/tests, and preparing this PR description. The 
approach follows maintainer discussion on #17786 (process kill alone is not 
always enough for streaming/cluster jobs) and SeaTunnel's documented 
`seatunnel.sh -can <jobId>` command.
   
   ## Purpose of the pull request
   
   Fixes #17786 (scoped to SeaTunnel Engine / Zeta)
   
   Stopping a SeaTunnel streaming task from DolphinScheduler can leave the 
engine job RUNNING: DS only killed the local client process, and 
`getApplicationIds()` always returned empty, so no engine-level cancel happened.
   
   ## Brief change log
   
   - Parse SeaTunnel Engine job id from worker logs (`job id: <id>`)
   - On cancel for `SeatunnelEngineTask`: keep process-tree kill, then run 
`/bin/seatunnel.sh -can <jobId>`
   - Add unit tests for job-id log parsing
   
   Out of scope for this PR: Flink/Spark Yarn cancel refactor, full status 
polling product, REST cluster stop (#15738).
   
   ## Verify this pull request
   
   This change added tests and can be verified as follows:
   
   - Added `SeatunnelJobIdParseTest` for SeaTunnel submit log lines
   - Manual: run a Zeta streaming job, Stop in DS UI, confirm SeaTunnel job 
becomes CANCELED (`seatunnel.sh -l` / SeaTunnel UI)
   
   ## Pull Request Notice
   [Pull Request 
Notice](https://github.com/apache/dolphinscheduler/blob/dev/docs/docs/en/contribute/join/pull-request.md)


-- 
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]

Reply via email to