SEPURI-SAI-KRISHNA commented on PR #12618: URL: https://github.com/apache/seatunnel/pull/12618#issuecomment-5981261815
Thanks @DanielLeens. Tracking issue drafted per your scope; I will open it separately so the guard stays out of this PR. On the red Build: I pulled the three failing job logs for run `37140738764` (head `48599a8f6`) and every failure is an already-tracked issue, none of them reachable from this diff. | job | failing test | tracked as | | --- | --- | --- | | `transform-v2-it-part-1 (11)` | `TestFilterRowKindIT.testFilterRowKindMultiTable` | #12116 | | `engine-v2-it (11)` | `BackpressureSlowSinkIT.testCheckpointsKeepCompletingUnderSustainedBackpressure` | #12313 / #12316 | | `paimon-connector-it (8)` | `PaimonWithS3IT.privilegeEnabledPaimonSourceUnAuthorized` | #11679 | A few details worth stating rather than asserting "flaky": - I first suspected timeouts, because the three jobs ran 97, 90 and 60 minutes. That is wrong: `backend.yml` on `dev` sets `timeout-minutes` of 135, 150 and 180 for those jobs respectively, so none of them hit its limit. They are real test failures. - `TestFilterRowKindIT` is the one that matters here since it is in this module. #12116 names that exact test and config (`filter_row_kind_exclude_insert_multi_table.conf`), reports the identical `expected: <0> but was: <1>` exit-code signature, states it reproduces only on the Flink legs (this was container `[2]`), and lists three unrelated PRs hitting it. Root cause there is the JVM-wide static counters in `AssertSinkWriter`, not the transform. - The paimon assertion message cites its own tracking issue in the failure text: "see https://github.com/apache/seatunnel/issues/11679";. - The run's overall conclusion is `cancelled`, so the aggregate check is red partly for that reason too. On whether this diff could cause any of them: every `src/main` change in this PR is `X.toUpperCase()` to `X.toUpperCase(Locale.ROOT)` or the `toLowerCase` equivalent, plus the `java.util.Locale` import. 11 call sites, no other statement changed. On a runner whose default locale is not Turkish, Azerbaijani or Lithuanian, those calls return identical values, so the change is a no-op at runtime on CI and cannot affect a row count, a checkpoint or an S3 privilege check. Happy to rerun the three jobs if you would rather see the aggregate green, though #12116 in particular will keep recurring on the Flink leg until it is fixed. -- 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]
