DanielLeens commented on PR #12328: URL: https://github.com/apache/seatunnel/pull/12328#issuecomment-5770873698
@davidzollo Thanks for the follow-up review. To make sure this doesn't stay stuck on a misunderstanding: my 2026-09-16 comment wasn't disagreeing with you in principle, just pointing out that "default" has two different meanings here, and they lead to different fixes. - If "default" means *the documented example's `map: engine*: map-store: enabled: true`* — that's already effectively `true` for `engine_finishedJobMetrics` today via wildcard inheritance, since there's no exact-match override. Setting it to `true` explicitly would be a no-op; it doesn't change today's behavior at all, and doesn't touch the unbounded-growth problem in #12022. What actually fixes that is the opt-out this PR adds (`engine_finishedJobMetrics: map-store: enabled: false`). - If instead you meant "the product's *actual* code-level default should be `false` for this map regardless of what a user's `hazelcast.yaml` says" — I agree that's the more robust fix, and it's exactly the follow-up I flagged in my original review: give `engine_finishedJobMetrics` the same hardcoded `NoOpMapStorage` guard in `FileMapStore.init()` that `engine_runningJobMetrics` already has (`seatunnel-engine-server/src/main/java/org/apache/seatunnel/engine/server/persistence/FileMapStore.java:46-53`). That's a behavior/data-retention change to production code, not a docs change, so I'd still keep it out of this PR and do it as its own scoped PR with its own tests. So: if your CHANGES_REQUESTED is about the second reading, I don't think it should block this docs fix — it's a separate, larger change that deserves its own review. If it's about the first reading, I don't think there's anything left to change, since the value is already `true` by inheritance today and this PR is what actually closes the gap. @albgen, no action needed on your side unless @davidzollo has a different, more specific ask than what's outlined above — happy to help scope the code-level follow-up if either of you wants to open it. -- 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]
