SEZ9 commented on PR #11677:
URL: https://github.com/apache/seatunnel/pull/11677#issuecomment-6105276196

   Thanks for the follow-up. I'd like to re-check my earlier points against the 
current head `8b430e980` rather than rely on summaries, so a few pointers would 
help:
   
   **Javadoc / contract scope (F1, F2)** — Scoping the `SourceSplitEnumerator` 
Javadoc back to what the engine actually guarantees, and documenting the 
stricter re-dispatch behavior only on `IncrementalSourceEnumerator`, is the 
approach I'd prefer. Could you point me to the relevant hunks in `8b430e980` so 
I can confirm the wording no longer implies a new obligation on other 
connectors' `addSplitsBack()`?
   
   **`if (running)` gate and pre-run `addSplitsBack()` (F3, F4)** — The 
ordering-invariant comments plus the `IncrementalSourceEnumeratorTest` case 
covering `addSplitsBack()` → `run()` sound like they address both the 
documentation and coverage gaps. Could you confirm that test asserts the 
restored splits are dispatched exactly once (no double assignment, no drop)?
   
   **E2E points (F5–F8)** — I couldn't tell from the discussion whether these 
were changed. Could you confirm, for `AbstractMysqlCDCITBase`:
   1. F5: whether the 30-second wait for the injected sink failure to surface 
in the logs was aligned with the 2-minute budgets used elsewhere, or why 30s is 
safely sufficient in CI.
   2. F6: whether the `DROP TRIGGER` in the `finally` block is guaranteed to 
take effect before the restarted job can replay the failing insert.
   3. F7/F8: whether `getServerLogs().substring(logOffset)` is guarded against 
the log string being shorter than `logOffset` (e.g. clamp the offset or catch 
and retry inside the awaitility lambda).
   
   If those are already handled in `8b430e980`, a quick pointer is all I need 
and I'm happy to approve.
   
   <!-- streview-comment:1673 -->


-- 
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