DanielLeens commented on PR #12165:
URL: https://github.com/apache/seatunnel/pull/12165#issuecomment-5812804131
@Vivek1106-04 Thanks for the quick turnaround on both items.
**1. PR description wording.** Confirmed — this closes Issue 4 from my last
review.
**2. CI rerun on this exact head (`c773cc091`).** I re-checked the rerun
independently rather than taking the summary on faith (fork run `35729456152`):
- Both Windows `unit-test` legs are now **green**. That matters beyond just
"passing": on the previous run they were red on the unrelated
`PayPalClientTest` flake before the reactor ever reached
`seatunnel-engine-server`, so this PR's own new/changed engine tests had never
actually been exercised on Windows. They have now, and they pass.
- `unit-test (8/11, ubuntu-latest)` and `engine-v2-it (11, ubuntu-latest)`:
green.
- `engine-v2-it (8, ubuntu-latest)`: still red, but on a **different**
failure than the one I traced last round. This time it's
`SplitClusterFaultToleranceIT#testStreamJobCancelResolvesWhenWorkerCrashesBeforeCancelAck`
(`Errors: 1` in the job log), not `BackpressureSlowSinkIT`. That's the
pre-existing `dev`-level race already tracked as #12353 (fix #12311, open and
unreviewed since 09-15) — same class of thing as before, just a different flaky
test in the same IT job, and it doesn't trace to anything in this diff.
- `jdbc-connectors-it-part-1 (11)` shows cancelled, not failed — no signal
either way, and that module isn't touched here.
So this clears the bar I set in my last review ("a clean run, or an explicit
per-job attribution on this exact head") — every red job on this run now
attributes to a known, pre-existing `dev` flake unrelated to this PR's code,
and the two legs that previously never even reached this PR's tests (Windows)
are now green on them. I don't have a new code-side concern to add on top of my
last review.
**3. #12441 handoff.** Agreed, that's a reasonable sequencing call — no
objection to landing the same-coordinator part there first and adding the
cross-pipeline regression once this PR is in.
Nothing further from me at this point.
--
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]
