hapostgres / hapostgres/pg_auto_failover
Improve pg_autoctl shutdown sequence
Open
enhancement
Size:M
user experience
- Dominant language
- C
- Stars
- 1.4k
- Forks
- 142
- Avg merge
- 5h 8m
- Merged PRs (30d)
- 1
Description
Both systemd and Kubernetes use a shutdown sequence that sends a SIGTERM signal then later a stronger one. See https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#pod-termination for instance, where the default is to provide pods with a 30s grade period.
At the moment pg_autoctl is not smart about the grace period. It could be made to switch the current node to MAINTENANCE mode and then have a way to automatically re-join the cluster at the next start-up.
Contributor guide
Assessment
This issue has not been assessed yet.