DanielLeens commented on PR #11206: URL: https://github.com/apache/seatunnel/pull/11206#issuecomment-5350357361
Follow-up self-check, ~12 hours after my last comment on this same head (`60d4e4155519`, unchanged — no new commit, no new discussion since). Re-deriving the full analysis again against a byte-identical diff wouldn't add anything on top of that comment, so this is a live re-verification of the two things most likely to have moved: CI and merge state. - **CI** — still `Build: fail` on this SHA, consistent with the deterministic (10/10) `testMysqlCdcByWildcardsConfigWithNewlyAddedTable` failure documented in my last comment (source-side `"Registered newly added CDC table"` and sink-side `"Registered runtime sink writer for newly created table"` logs still never appear; `CreateTableEvent` still never appears anywhere in the job log for this head). - **Merge state** — still `mergeStateStatus=DIRTY` / `mergeable=CONFLICTING` against `dev`. Divergence has narrowed slightly (`ahead_by=14`, `behind_by=18`, vs. `behind_by` in the high-30s/40s range on some earlier rounds of other PRs in this same author's queue), but the conflict itself is unresolved. - Nothing in this window changes the three High blockers from my last comment: (1) the feature's own e2e test has never been observed to pass, (2) a binlog-registered table isn't checkpointed and would be silently lost on restart, (3) no capture-pattern filter is applied before registering a DDL-discovered table, contradicting this PR's own documentation. This remains the live technical conclusion for this diff — I'm not aware of anything in the intervening ~12 hours that would change it. ### Conclusion: Not recommended for merge Unchanged from my last comment: root-cause and fix the e2e test failure (Issue 1) before investing further review cycles in the durability (Issue 2) and scoping (Issue 3) gaps, then resolve the `dev` conflict independently of the above. As before, this is posted as a plain comment rather than a formal review since GitHub does not allow self-approval; it carries no approval weight and merge still requires review from another committer. -- 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]
