nzw921rx opened a new pull request, #11542: URL: https://github.com/apache/seatunnel/pull/11542
## Purpose of this pull request Stabilize two flaky Java 8 unit tests observed in consecutive CI attempts of #11527. ## Changes - Count assigned slots for the current job in `JobMasterTest#testCloseIdleTask` instead of using the worker-wide slot count. This prevents asynchronously released slots from other jobs from causing an exact-count timeout. - Use force cancellation in `SeaTunnelEngineClusterRoleTest#testWorkerIsFirstMemberThenGetJobDetailStatus`. The test verifies job queries after a master handoff, so it should not block on graceful checkpoint teardown on the promoted master. ## CI comparison - Attempt 1 timed out in `JobMasterTest#testCloseIdleTask`. - Attempt 2 timed out in `SeaTunnelEngineClusterRoleTest#testWorkerIsFirstMemberThenGetJobDetailStatus` while waiting for `CancelJobOperation`. The different failures across attempts, together with both tests passing in isolation before the change, indicate independent test flakiness rather than a failure caused by the MongoDB connector changes in #11527. ## Validation - `./mvnw spotless:apply` - `./mvnw -pl seatunnel-engine/seatunnel-engine-server -DskipIT -DskipUT=false -Dtest=JobMasterTest#testCloseIdleTask -DfailIfNoTests=false test` - `./mvnw -pl seatunnel-engine/seatunnel-engine-client -DskipIT -DskipUT=false -Dtest=SeaTunnelEngineClusterRoleTest#testWorkerIsFirstMemberThenGetJobDetailStatus -DfailIfNoTests=false test` Both focused tests passed on Java 8. -- 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]
