Rancher Backups and Disaster Recovery for RKE1

Sun, 30 Jun 2024 23:18:00 -0500

Eventually, something will happen that requires restoring a Kubernetes cluster. That’s not a question of if but when.

In the case of Kubernetes, this means regularly backing up the etcd datastore.

How regular? The choice is yours. It is up to you how much data you are willing to lose.

For example, a highly active cluster may take snapshots every fifteen minutes or even every five minutes. In the case of a cluster that does not receive a lot of activity, a snapshot of the cluster might be taken daily.

You can never be fired for having too many backups.

As a default, RKE takes a snapshot every six hours and keeps it for one day, but you can change the settings in the cluster.yaml.

Any time you wish, you can run a snapshot by executing the command rke etcd snapshot-save and passing it the name of the backup and the location of cluster.yml. This command will write a snapshot of all the etcd hosts to /opt/rke/etcd-snapshots. It is possible to mount an NFS volume there, or you can configure RKE to copy the snapshot to S3 to restore it.

RKE1 on Support Tools

Rancher Backups and Disaster Recovery for RKE1