Jackie-Jiang opened a new pull request, #19679:
URL: https://github.com/apache/pinot/pull/19679

   ## Summary
   
   A waiting commit can acquire an empty queue after another leader removes its 
entry, but before that leader signals it. The early return used 
`_updatedIdealState` and skipped `_exception`, so a failed updater could appear 
successful to its owner. Route that path through the common outcome check.
   
   Add a deterministic regression test that records a failed batch for an 
already removed entry, then lets its owner resume.
   
   ## Failure
   
   The [JDK 25 unit-test 
job](https://github.com/apache/pinot/actions/runs/36202559517/job/108292131826?pr=19678)
 for #19678 failed in 
`IdealStateGroupCommitTest.testMultipleStuckAndFreshConsistency` with `Stuck 
owner 2 should have thrown`. The log also shows `PermanentUpdaterException: 
stuck-2` in a different worker, consistent with the owner bypassing its 
recorded exception.
   
   ## Validation
   
   The new test asserts that the owner receives the recorded failure in this 
exact interleaving. GitHub Actions will run the affected modules on this PR.


-- 
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]

Reply via email to