SEZ9 commented on issue #12617:
URL: https://github.com/apache/seatunnel/issues/12617#issuecomment-6007925712

   Thanks for the pointers, and sorry for the duplicated asks — taking your 
three in the same order.
   
   **(c) #12620** — understood, that covers what I was after: the 24-site 
baseline on `af80a6704` (13 in `sql/` via #12495, 11 via #12618), guard kept 
out of #12618, and no Checkstyle/forbidden-apis dependency until the scope is 
agreed. Nothing further needed there; just hold implementation until #12495 and 
#12618 have both landed so the guard runs against the post-fix module.
   
   **(b) Reachable-vs-hardening split** — the five/six breakdown in the #12618 
description is exactly the note I wanted recorded. Having 
`ModalityType:60/78/102`, `PayloadFormat:40` and `ModelInvocationCacheKey:144` 
called out as the reachable set, with `BuiltinFunctions:54` and the other five 
explicitly labelled consistency changes, makes the diff easy to read against 
the four tests. No change requested.
   
   **(a) #12618 Build** — your triage of run `37140738764` on `48599a8f6` is 
convincing: `TestFilterRowKindIT.testFilterRowKindMultiTable` (#12116), 
`BackpressureSlowSinkIT` (#12313 / #12316) and 
`PaimonWithS3IT.privilegeEnabledPaimonSourceUnAuthorized` (#11679) are all 
tracked, and I agree that a `toUpperCase()` → `toUpperCase(Locale.ROOT)` 
substitution plus an import cannot move a row count, a checkpoint or an S3 
privilege check on a non-Turkish runner. Thanks also for correcting the timeout 
reading against the `backend.yml` limits rather than letting it stand.
   
   Two small asks to close this out:
   
   1. Please do rerun the three failed jobs once. I don't need the aggregate to 
be green to proceed, but a second run where only the known-flaky legs fail (or 
none do) gives reviewers something to point at instead of the triage comment, 
and if `transform-v2-it-part-1 (11)` fails again on #12116 we simply note it.
   2. Drop a one-line cross-link from #12618 back to #12620 so the "guard is 
intentionally not here" decision is discoverable from the PR itself.
   
   Otherwise I have nothing outstanding on this issue; it stays open until 
#12618 merges into `dev`.
   
   <!-- streview-comment:1540 -->


-- 
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]

Reply via email to