Using OCI Database with PostgreSQL Backups in Disaster Recovery

Backups can be copied and distributed to another region. You can use a backup that's been copied to a second region to provision a new database system if the primary database system is unavailable. The disaster recovery process using a backup follows:

1. Copy a database system backup. You can copy backups in a few ways, depending on whether the database system already exists:

   1. Enable backup copies when creating the database system.
   2. Enable backup copies for an existing database system by editing its management policy.
   3. Copy an existing backup to another region.
   Note
   
   We recommend using a management policy that automatically creates and copies backups.
2. If the primary region is unavailable, use the backup that was copied to another region to create a new database system in that region. If the primary database system used a custom configuration, you might need to create a similar configuration in the backup region as well. When creating the database system, enable automatic backups so that you can restore back to the primary region later.
3. After the new database system is available, redirect applications to use its endpoint.
4. When the primary region becomes available, copy a backup to the original region.
5. Restore the original database system using the backup you copied to its region.
6. After the original database system is restored, redirect applications to use its endpoint.

Depending on the schedule of the management policy, RPO could vary from a few minutes to greater than 24 hours. Using the preceding process, RTO could be a matter of minutes.