Rangsh commented on PR #12081: URL: https://github.com/apache/seatunnel/pull/12081#issuecomment-5570535004
@DanielLeens thanks for the full re-review of `26ce5cb` — and for confirming the Related-to-#12058 framing. Agreed on the split: this PR is correctness-first; #12058 stays open for the CV investigation on its own timeline. Pushed `f08bdf0ac` addressing your blockers and the non-blocking notes: ### Blockers 1. **Issue 1 (WAL worker survivability)** — added `WALWorkHandlerSurvivabilityTest`: first `writer.write()` throws a non-`IOException` (`IllegalStateException`); `onEvent` returns normally with `done(false)`; a subsequent APPEND still completes with `done(true)`. Proves the sole consumer is not killed and keeps processing. 2. **Issue 2 (shared batch deadline)** — added `IMapFileStorageBatchDeadlineTest`: 5 never-completed futures with `writDataTimeoutMilliseconds=200ms`; asserts all keys fail and wall time stays near **1× timeout**, well under **N× timeout**. ### Non-blocking - **Issue 3 (`deleteAll` double-publish)** — agreed out of scope; will file a separate follow-up issue/PR. - **Issue 4 (Mockito dep)** — kept the explicit module `mockito-junit-jupiter` entry with a clearer comment (explicit opt-in / classpath visibility; also inherited from root). - **Issue 5 (`CLOSED` unguarded)** — added a one-line note that `CLOSED`/`close()` is intentionally shutdown-path-only and does not threaten steady-state APPEND persistence. Happy to re-review whenever you have a chance. -- 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]
