SEPURI-SAI-KRISHNA commented on PR #12899: URL: https://github.com/apache/seatunnel/pull/12899#issuecomment-6359637649
CI result on head `e72b56381`. **Build is red, but your Issue 1 is cleared.** 96 jobs: 84 success, 11 skipped, **1 failure**. | Lane | Prev `aaa915e28` | Now `e72b56381` | |---|---|---| | `unit-test` x4 | all **failed** | all **pass** | | `transform-v2-it` x4 | pass | pass | | `engine-v2-it (11)` | pass | pass | | `engine-v2-it (17)` | **pass** | **fail** | | `mysql-cdc-it` x2 | cancelled | pass | The `JdbcOptionRuleCompletenessTest` break you traced to #12874 is gone, confirming #12938 carried the cure after #12875 was closed unmerged. **The single failure**, in [`engine-v2-it (17)`](https://github.com/SEPURI-SAI-KRISHNA/seatunnel/actions/runs/39808422784/job/119929294401), read from the test-reports artifact rather than the job log, which truncates: ``` BackpressureSlowSinkIT#testCheckpointsKeepCompletingUnderSustainedBackpressure expected at least 3 additional checkpoints to complete during the 90s sustained backpressure window, only observed 1 (samples=[1, 1, 1, 1, 1, 1, 2, 2, 2]) at BackpressureSlowSinkIT.java:247 ``` Artifact totals for that job: 221 tests, 1 failure, 7 skipped. **Not attributable to this diff.** This PR touches `SystemFunction.java`, two of its test classes and four docs, nothing under `seatunnel-engine` or `seatunnel-e2e`. The lanes exercising the change, four `unit-test` legs and both `transform-v2-it` parts, are green. The same lane passed on the previous head here and `engine-v2-it (11)` passes on this one, so it is JDK-17 only. Your #12313 names this exact test and assertion shape, with #12316 for the lock handoff; both still open. **A sync is worth doing.** The branch is behind 11 / ahead 3, and four of those 11 touch engine checkpoint code: - `969a46d29` adds `receivedSourceEvent` to `SourceFlowLifeCycle`. Additive; `triggerBarrier` and the lock are unchanged, though line numbers shift from those #12313 cites. - `df94fb07d` wraps the periodic trigger so a thrown trigger fails the coordinator instead of silently dropping the scheduled chain. - `d33bd82a4` master-failover restore; `5a5041364` a REST field. None changes the lock handoff, so I will not claim a sync fixes this. But `df94fb07d` would surface a swallowed trigger exception if one were involved, so re-syncing and rerunning that lane gives real information rather than a blind roll. Happy to do that if you agree. **Issue 2** is on this head: `SystemFunctionTest.java` has one occurrence of the qualified name, the import on line 28, and no inline usages. -- 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]
