DanielLeens commented on PR #11512: URL: https://github.com/apache/seatunnel/pull/11512#issuecomment-5726069464
Thanks @goutamadwant for rechecking F4/F6, and @SEZ9 for confirming both at `d8cb0cbb7d3`. On your first ask -- whether the shared integration failures are unrelated to this PR's changes -- I pulled the actual job logs from the fork's completed `Build` run (`goutamadwant/seatunnel` run `34918501024`, head `d8cb0cbb7d39c`) rather than going from job names alone: - `all-connectors-it-2` (both JDK legs): connector-IT infra, outside the `seatunnel-api`/`seatunnel-engine-server` files this PR touches. - `unit-test (11, ubuntu-latest)`: `RestApiSubmitJobConfigShadeDecryptTest.setUp` fails with `IllegalStateException: Node failed to start!` -- this is the Hazelcast-discovery/fixture issue @goutamadwant already isolated and filed as #12338, not new. - `engine-v2-it (8, ubuntu-latest)`: `BackpressureSlowSinkIT.testCheckpointsKeepCompletingUnderSustainedBackpressure` -- the source checkpoint-lock-starvation flake tracked in #12316/#12313, both still open on `dev`. - `engine-v2-it (11, ubuntu-latest)`: `CheckpointCoordinatorFailoverIT.testStreamJobFailsAfterCheckpointTriggerDispatchFailure` and `SavepointBusySourceBarrierIT.testStopWithSavepointCompletesWhileSourceEmitsLargeSplit`, plus a `ClusterFaultToleranceIT` restore-in-all-node-down case. The last group is the one I can't fully close out yet: they exercise checkpoint-coordinator/failover paths directly, and this PR's diff (the CDC progress contract plus the `TaskExecutionService.reportCdcProgressToMaster` reporting hook) doesn't touch any checkpoint-coordinator code, so on file scope they look unrelated too -- but unlike the first three, they haven't been triaged against a clean `dev` run in this thread yet, so I'd treat "unrelated" as a plausible read rather than a confirmed one until a `dev` baseline or a rerun shows the same failures there. On your second ask: agreed, that's the right next step. Once you rebase onto current `dev` and get a fresh run, I'll do a full pass, and if the aggregate `Build` comes back clean (or only with the tracked items above), that closes the CI gate from my side. -- 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]
