Skip to main content

Your submission was sent successfully! Close

Thank you for signing up for our newsletter!
In these regular emails you will find the latest updates from Canonical and upcoming events where you can meet our team.Close

Thank you for contacting our team. We will be in touch shortly.Close

Remove or recover a cluster

This guide will cover how to manage clusters using an example PostgreSQL deployment with two servers: one in Rome and one in Lisbon.

Prerequisites

Summary


Switchover

If the primary cluster fails or is removed, it is necessary to appoint a new cluster as primary.

To switchover and promote lisbon to primary, one would run the command:

juju run -m lisbon db2/leader promote-to-primary

Detach a cluster

Clusters in an async replica set can be detached. The detached cluster can then be either removed or reused.

Assuming lisbon is the current primary, one would run the following command to detach rome:

juju remove-relation -m lisbon replication-offer db2:replication

The command above will move the rome cluster into a detached state (blocked) keeping all the data in place.

Reuse a detached cluster

The following command creates a new cluster in the replica set from the detached rome cluster, keeping its existing data in use:

juju run -m rome db1/leader promote-to-primary

Remove a detached cluster

The following command removes the detached rome cluster and destroys its stored data with the optional --destroy-storage flag:

juju remove-application -m rome db1 --destroy-storage

Recover a cluster

If the integration between clusters was removed and one side went into a blocked state, integrate both clusters again and call the promote-cluster action to restore async replication - similar to the “Reuse a detached cluster” step above.

If the cluster group lost a member entirely (e.g. rome is suddenly no longer available to the cluster group originally consisting of rome and lisbon), deploy a new postgresql application and set up async replication. The data will be copied automatically after the promote-cluster action is called, and the new cluster will join the cluster group.

Last updated a month ago. Help improve this document in the forum.