SEZ9 commented on issue #12344: URL: https://github.com/apache/seatunnel/issues/12344#issuecomment-5723706105
@davidzollo — pinging you specifically because this one needs a committer's decision and I think you already have a stake in it rather than because I want to hurry anyone. Your #12314 fails `all-connectors-it-2` on both JDK 8 and JDK 11, the same job and the same test as the pull requests I mentioned above: ``` FAIL Run / all-connectors-it-2 (8, ubuntu-latest) FAIL Run / all-connectors-it-2 (11, ubuntu-latest) ``` To be fair to the facts: #12314 has other failing jobs too (`unit-test` on four platforms, `engine-v2-it`, `all-connectors-it-7`, `rocketmq-connector-it`), so this issue is not the only thing standing between it and a green `Build` — but it is one of them, and it is the one no amount of work on the PR can clear. The decision I'm asking for is narrow. `testAddFieldWithRestore` has failed 40 of 40 real executions and has never been observed passing, so it is not flakiness that reruns absorb. Either: 1. someone who knows the Opengauss CDC restore path fixes it — it may well be reporting a genuine bug, which is exactly why I don't want to silence it on my own judgement; or 2. it gets quarantined with `@Disabled` and a link back to this issue, so `all-connectors-it-2` stops gating every Zeta and core-API pull request while the fix is worked out. If (2) is the call, say the word and I'll open that PR — it's a two-line mechanical change and I can have it up within the hour. If (1), I'll stop bringing it up and wait; I just didn't want the choice to go unmade by default, since #12311 and #12313 — the fixes for the two *other* tests currently failing in `engine-v2-it` — are blocked behind this same job. -- 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]
