Pause/resume and standby mode for a PostgreSQL cluster

Pause and resume the cluster

You can temporarily shut down your PostgreSQL cluster and bring it back later without losing data or configuration. You may want to pause the cluster for maintenance tasks, emergency manual intervention or debugging.

When paused, all changes to the cluster’s current state are suspended and no statuses other than the “Progressing” condition are updated until you resume the reconciliation.

How to pause

Set the spec.pause option to true in your deploy/cr.yaml Custom Resource:

spec:
  pause: true
  # ... rest of your spec

Apply the change:

kubectl apply -f deploy/cr.yaml

The Operator will gracefully stop the cluster (primary, replicas, pgBackRest, and related jobs).

How to resume

Set spec.pause back to false in the same Custom Resource and apply:

spec:
  pause: false
  # ... rest of your spec

kubectl apply -f deploy/cr.yaml

The Operator will start the cluster again using the existing data volumes.

Troubleshooting

The Operator does not pause the cluster

If a backup job is running, the Operator will not pause the cluster and will log a warning. Remove a running backup job so you can pause:

kubectl delete job -l postgres-operator.crunchydata.com/pgbackrest-backup -n <namespace>

Then retry pausing the cluster.

Standby mode

Standby PostgreSQL clusters provide a continuously replicated copy of your primary cluster, forming the backbone of high‑availability and disaster‑recovery strategies. They stay in sync through streaming replication, enabling you to quickly promote a standby if the primary becomes unavailable. Standby clusters can also run in separate regions or environments, helping you maintain business continuity during outages.

The standby mode for a cluster is controlled with the spec.standby.enabled option plus the spec.standby.repoName and/or spec.standby.host and spec.standby.port options in the Custom Resource. What options to specify depends on the standby cluster type.

Read more about the supported types of standby clusters and their setup in the Deploy a standby cluster for Disaster Recovery documentation.

---

Last update: February 12, 2026
Created: February 12, 2026