There are two ways
- manual (snapshots or via aws backups)
- automated (1 - 35 days)
Recovery
- Automated
- AWS Console → RDS → “Restore to point in time”.
- Creates a new DB instance from the backup + logs.
- Manual
-
Console → RDS → Snapshots → Select → “Restore Snapshot”.
-
Creates a new DB instance from that snapshot.
-
Backup console → Recovery points → “Restore”
-
Also creates a new RDS instance.
-
When
- Single AZ
- Backup takes I/O
- PITR and snapshots
- Muliti AZ
- Takes from standby instance, no I/O effects
- PIRT and Snapshots
- RR
- Replicas aren’t used for backups, still taken from primary instance, partly because replication is async
- Why would you back up a back up (kinda)
- If you want, promote to master and take it from there