SEZ9 commented on PR #11538:
URL: https://github.com/apache/seatunnel/pull/11538#issuecomment-5230720145

   Thanks @goutamadwant for the thorough follow-ups — this is shaping up well.
   
   Confirming the points that now look closed to me:
   
   - **Parallelism**: forcing source and transform parallelism to 1 in 
`MultipleTableJobConfigParser`, with the two-source parallelism-3 regression 
test, resolves my determinism concern about multiple sample collectors.
   - **CLI contract**: moving sample mode onto the validated CLI path (instead 
of env options), rejecting `--restore-with-checkpoint` and other conflicting 
options, and making row output opt-in matches what I asked for.
   - **Semantics**: the clarified `--sample-limit` wording ("limits rows 
forwarded into the sample pipeline; a source may finish its active poll/batch 
before stopping") is honest and useful — good to have that documented rather 
than implied.
   - **Schema logging**: logging only the table path and physical row schema, 
with the password-option regression test proving `CatalogTable` options are 
excluded, addresses the sensitive-data-in-logs concern.
   
   Remaining asks before I approve:
   
   1. **CI / merge gate**: I agree the Java 8 RocketMQ (20 vs 25 messages) and 
Doris (`UnknownHostException: doris_e2e`) failures look like unrelated E2E 
flakiness, especially with the Java 11 twins and both engine V2 integration 
jobs green. That said, the branch is still diverged from `dev` — please 
rebase/merge latest `dev` and rerun the failing jobs so we have a green (or 
clearly-flaky-only) signal on the merged base.
   2. **Minor cleanups from my earlier pass** — please confirm whether these 
landed in the recent commits, as I didn't see them called out explicitly:
      - the stale en `Usage` block relative to the new `--dry-run` description,
      - the duplicated `Common.setDeployMode` call,
      - `ParameterException` thrown from `execute()` outside parse time,
      - Javadoc on the new public members.
   
   If those four are done (or you point me to the commits), I'll do a final 
head-to-head diff read after the CI rerun. Thanks again for the disciplined, 
test-backed iterations!
   
   <!-- streview-comment:98 -->


-- 
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