SEZ9 commented on PR #11721: URL: https://github.com/apache/seatunnel/pull/11721#issuecomment-5421605382
Thanks @SEPURI-SAI-KRISHNA for the detailed follow-up — this answers exactly what I flagged as UNCERTAIN on F1, so let me close that out directly. **F1 — resolved.** Your table confirms what I needed: zero occurrences of `extractPrimaryKeyIfPresent` and `% replicaNum` in both the en and zh multi-table docs at `aea9854a1`, the routing expression at `:311`/`:405` matching `MultiTableSinkWriter.java:622` token for token, and the key extraction line using `element.getField(primaryKey.get())`. The disclaimer at `:249-253` labeling the listing as simplified and "not a copy of the source" also satisfies the second remedy path I offered — so both, not just one, are covered. @DanielLeens independently re-verifying all of this at `aea9854a1bb1` in a fresh worktree settles it. F1 is closed on my side. **On the zh `\bmod replicaNum` question** — thanks for surfacing it rather than glossing over it. Your trace through `MultiTableSink.java:147` passing `replicaNum` as `queueSize`, plus the one-queue-per-`queueSize` construction, shows `blockingQueues.size() == replicaNum` by construction, and DanielLeens confirmed the same chain independently. Given the zh section is framed around replica tuning, the current wording is fine — no change needed. If you'd rather switch it for cross-language symmetry anyway, I won't object, but it's not a condition for merge. **F2 — deferral agreed.** Keeping the shared non-negative-mod helper out of a one-line bug fix is the right call given the sibling call sites you found span multiple modules. The one concrete remaining ask: please file the follow-up issue as promised, including the `Integer.MIN_VALUE` unit test, and drop the link here so we have the paper trail. Also thanks @DanielLeens for the CI triage on run `32746106023` — the stuck-runner read on the cancelled kudu job sounds right, and it's untouched by this diff regardless. With F1 confirmed and the F2 follow-up as the only outstanding item, nothing blocks this from my side either. <!-- streview-comment:574 --> -- 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]
