For AI agents: The documentation index is at https://docs.digitalocean.com/llms.txt. Markdown versions of pages use the same URL with index.html.md in place of the HTML page (for example, append index.html.md to the directory path instead of opening the HTML document).

DigitalOcean Managed Databases automatically recover from hardware and software failures by replacing degraded nodes with new ones that resume from the point of failure.

To recover from accidental data loss, such as when a database administrator or application deletes data, manually restore a PostgreSQL database cluster from a backup.

Restoring creates a new copy of the cluster’s primary node. A restore operation creates a new primary node because restoring into the existing primary node would create multiple database timelines, making recovery history more difficult to track and manage. Creating a new primary node preserves a single linear recovery history.

PostgreSQL cluster backups run automatically once per day and are retained for seven days. Backups don’t cause downtime or reduce performance. DigitalOcean automatically sets the backup time, which can’t be modified.

Restore a PostgreSQL Cluster from Backups

To restore from a backup, from the Databases page, click the PostgreSQL database you want to restore. On its Overview page, click the Actions dropdown menu, and then select Restore from backup.

If you can’t select Restore from backup, your account or team may have reached its database cluster limit. To proceed, delete an existing cluster from the Databases page, or contact support to request a limit increase.

In the Create a new cluster from a backup window, choose whether you want to restore to the latest transaction available or choose a point in time.

Afterwards, choose a name for the new database. By default, the name is the original database’s name appended with the date of the backup and the word “backup”, like originalname-aug-13-backup.

When finished, click Restore to New Cluster to begin the restoration. The time it takes to create the new cluster depends on the amount of data in the original.

When the restoration completes, you can add standby nodes and read-only nodes as needed.