SEZ9 commented on PR #12710: URL: https://github.com/apache/seatunnel/pull/12710#issuecomment-6207355847
Thanks for keeping the branch current. Between `d6f7513761` and `bd58ded09c` the PR-owned files appear unchanged, so I only have a couple of follow-up questions on my earlier points rather than anything new: 1. **Master/worker Deployment shutdown behaviour** – With `exec java` the JVM now receives SIGTERM and the Hazelcast shutdown hook runs, which is good. Could you point me to where the chart sets `terminationGracePeriodSeconds`, and confirm whether a `hazelcast.shutdownhook.policy` is configured? I want to verify whether the resulting stop is Hazelcast `TERMINATE` (member leaves promptly, tasks recover from the last checkpoint) or a graceful one, and make sure the description and docs match whichever it is. Whatever is decided should apply consistently to both the master and worker Deployments. 2. **Kubernetes graceful-shutdown docs** – Could you link the en/zh doc changes that describe the new SIGTERM path, and confirm whether the `preStop` + `stop-seatunnel-cluster.sh` section was updated? I'd like to check that they state clearly what the new path does and does not guarantee. The core `exec` change in `seatunnel-cluster.sh` and the exec-form `command` in the Helm templates look correct to me. Happy to take another pass once the above is clarified. <!-- streview-comment:1789 --> -- 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]
