ricky2129 commented on PR #12710: URL: https://github.com/apache/seatunnel/pull/12710#issuecomment-6107040498
@SEZ9 one related question before the next round. While checking the chart I noticed: - `values.yaml` has `master.readinessProbe` / `worker.readinessProbe` (`enabled: true` plus timings), but no template renders it, so pods never get a readiness probe. - There is no `startupProbe`, `lifecycle`/`preStop` passthrough or `PodDisruptionBudget` in the chart either. The hybrid/separated-mode docs show these only in hand-written StatefulSet manifests. Without a readiness probe, a rolling update can terminate the last old master as soon as the new pod is `Running`, before it has joined the cluster. Now that a stopped pod leaves in about a second instead of lingering for the grace period, that window matters more. Would you prefer this in the current PR, or as a separate follow-up PR? The follow-up would: - render `readinessProbe` and `startupProbe` like `livenessProbe`, with `tcpSocket` on `hazelcast-port` as the default; - add an optional PDB per role, off by default; - add a `lifecycle` passthrough. I'd lean towards a separate PR to keep this one focused on signal delivery, but happy either way. -- 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]
