SEZ9 commented on PR #12710:
URL: https://github.com/apache/seatunnel/pull/12710#issuecomment-6258991787

   Thanks for the follow-up on the shutdown and docs points. Based on the 
current state at `bd58ded09c` / `4667531b6f8a6e43a5f282a322e5df95a3f40e40`, 
here is where I land on the earlier findings:
   
   **PR12710-F1 / PR12710-F2 (shutdown behaviour, master and worker 
templates)** — `terminationGracePeriodSeconds` is now wired into both 
`deployment-seatunnel-master.yaml` and `deployment-seatunnel-worker.yaml` and 
driven from `values.yaml`, so the abrupt-teardown concern on the chart side is 
addressed. I'm fine with leaving `hazelcast.shutdownhook.policy` unset and 
keeping the default `TERMINATE` behaviour, as long as the description and docs 
are explicit that this is a fast, clean exit from the cluster with tasks 
recovering from the last checkpoint, not a graceful stop of running jobs. The 
thread indicates `hybrid-cluster-mode.md` and `separated-cluster-mode.md` 
already say this; please double-check the PR description uses the same wording.
   
   The `SeaTunnelServer.shutdown` terminate-flag part of F1 remains out of 
scope for this PR, which is acceptable, but I'd like it tracked: please open a 
small follow-up issue for it and link it from the PR description so it isn't 
lost.
   
   **PR12710-F3 (docs)** — Having the "Health Check and Shutdown" section lead 
with the exec-form `command` plus `terminationGracePeriodSeconds`, with 
`preStop` + `stop-seatunnel-cluster.sh` demoted to an optional tip for `/bin/sh 
-c` setups, is exactly what I was asking for. Considering this resolved.
   
   Remaining asks before merge:
   1. Tracking issue for the `SeaTunnelServer.shutdown` terminate-flag 
follow-up, linked in the description.
   2. A green `Build` run — it's currently failing on the fork; if the failure 
is the unrelated `SplitClusterFaultToleranceIT` flake, re-running only the 
failed jobs should be enough. Please ping once it's green.
   
   No further code changes requested from my side.
   
   <!-- streview-comment:1875 -->


-- 
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]

Reply via email to