Data Recovery Service automatic backups

By default, Cloudera Data Services on premises 1.5.4 and 1.5.4 SP1 versions enable Data Recovery Service automatic backups for the Cloudera Control Plane, Cloudera Data Engineering, and Cloudera Data Warehouse in the compute cluster of Embedded Container Service. The automatic backups are stored in the Longhorn in-cluster storage. You can also configure the external storage in Embedded Container Service, and then initiate the automatic back ups to it.

The following storage options are available to store the Data Recovery Service automatic backups in Embedded Container Service:

In-cluster storage

By default, Data Recovery Service automatic backups use Longhorn in-cluster storage. If necessary, you can configure the storage configuration settings in Longhorn by navigating to the > Cloudera Manager > Clusters [***CLUSTER NAME***] > Status > [***ECS CLUSTER NAME***] > Web UI > Storage UI page.

By default, Kubernetes initiates the first automatic backup within 30 minutes after the backup policy creation is complete, and then takes subsequent backups every hour.

You can change the backup retain count to take backups on an hourly, daily, or weekly basis and you can also disable the Data Recovery Service automatic backup functionality (set ENABLED to false) using the kubectl edit cj automatic-backup -n cdp-drs command. For more information about using this command in Data Recovery Service, see Initiating DRS automatic backups.

External storage

Embedded Container Service uses Longhorn as the underlying storage provisioner. In Longhorn, you can store snapshots externally using an S3 compatible storage such as Ozone or NFS v4. After you configure the external storage, edit the automatic-backup cron job to initiate the automatic backups.