markap14 commented on PR #11647:
URL: https://github.com/apache/nifi/pull/11647#issuecomment-5601856938

   [Cursor Grok 4.6] **macos-15 Java 21** failed on [run 
34343200146](https://github.com/apache/nifi/actions/runs/34343200146/job/102438584631)
 in 
`ClusteredConnectorTroubleshootingIT.testEnterAndExitTroubleshootingRestoresFlow`.
 ubuntu-24.04 Java 21/25 and macos-15 Java 25 on the same run passed, including 
the rest of that clustered troubleshooting class (17 other tests).
   
   ```
   java.lang.IllegalStateException: Connector failed to update
        at 
org.apache.nifi.tests.system.NiFiClientUtil.applyConnectorUpdate(NiFiClientUtil.java:511)
        at 
org.apache.nifi.tests.system.connectors.ConnectorTroubleshootingIT.testEnterAndExitTroubleshootingRestoresFlow(ConnectorTroubleshootingIT.java:149)
   ```
   
   Node 2 finished the local update (`UPDATING` → `UPDATED`), then 
`waitForState(..., UPDATED, allowable=UPDATING)` immediately saw cluster state 
`PREPARING_FOR_UPDATE` because node 1 had not finished starting the replicated 
apply yet. Node 1 then failed waiting for `UPDATING` after node 2 already 
aborted to `UPDATE_FAILED`. That wait does not treat `PREPARING_FOR_UPDATE` as 
in-progress, so a few milliseconds of cluster lag fails the update. 
Pre-existing clustered apply-update race, not the working-context lease. Reran 
the failed shard.


-- 
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