DanielLeens commented on PR #11639: URL: https://github.com/apache/seatunnel/pull/11639#issuecomment-5340762146
Re-checked this PR today. Head is still `afb0fada4b164062ef719f273caff9ad1197428d` — the same commit my full re-review at https://github.com/apache/seatunnel/pull/11639#pullrequestreview-4945111316 covered — and there has been no new commit and no new reply since then, so I did not publish a second full Steps-1-11 review against the same code version (that would just duplicate Issues 1-8 above against unchanged code). Instead, this is a re-verification round: - **CI**: still green, and it's literally the same run (`Build` / run `91769082988`, completed 2026-08-04) — nothing to re-diagnose. - **Merge state**: `mergeable: true` (no textual conflict), `mergeable_state: blocked` because the PR is still a draft with `reviewDecision: REVIEW_REQUIRED` — matches Issue 8. - **Upstream drift**: `dev` has moved further away since my last check (`behind_by` is now 85 vs. 60 three days ago). I walked every `dev` commit added in that window that touches `pom.xml`, `seatunnel-connectors-v2/connector-cdc/**`, or `seatunnel-dist/**` to see whether anything on `dev` now interacts with the shade-filter mechanism this PR depends on (the `connector-cdc` parent's `<filters combine.children="append">` block and the `connector-cdc-opengauss` manual restatement). Nothing does — the one CDC/shading-adjacent commit in that range, #11522 ("Stop unrelocated openGauss driver from hijacking PostgreSQL connections"), only touches `seatunnel-dist/pom.xml`'s driver packaging and a new `DriverJarPackagingTest`, not `connector-cdc/pom.xml` or `connector-cdc-opengauss/pom.xml`. So the shade-inheritance chain I traced on 08-16 is still accurate as-is; the growing `behind_by` is queue metadata, not a new blocker, and I'm not asking for a sync here since nothing about the curr ent CI failure needs it (CI isn't failing). - **Blockers unchanged**: since no commit has landed, Issue 2 (High — the third-party migration note in `docs/en` and `docs/zh` `incompatible-changes.md` omits the mandatory `io.debezium:debezium-core` shade exclusion, leaving out-of-tree CDC connectors exposed to a silent, order-dependent class-resolution race against the patched `ChangeEventQueue`) and Issue 1 (Medium — the same doc bullet incorrectly claims `connector-cdc-base` ships zero `io.debezium` classes, when it still ships five patched ones) are both still open at this head. That's the same class of risk as the Jackson-relocation regressions this project has hit before with cross-module shading changes: the mechanism itself is correctly implemented and verified for every in-tree connector, but the migration guidance for anyone building *outside* the tree is incomplete, which is exactly the audience least likely to notice before they hit an OOM in the snapshot phase. Conclusion is unchanged from the 08-16 review: **Ready to merge after fixes**, blocked on Issue 2 (High) and Issue 1 (Medium), with Issues 3-7 as non-blocking follow-ups and Issue 8 (draft status / rebase) still applying. I'll take another full pass as soon as a new commit lands. -- 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]
