Backup and restore
Managed Postgres ensures the safety and availability of your data through automated backups and point-in-time recovery. You can view your backup history and initiate restores from the Backups view of your instance.
Backups
Backup frequency
Managed Postgres takes daily full backups of your database. In addition to full backups, Write-Ahead Log (WAL) files are archived every 60 seconds or when 16 MB of WAL data accumulates, whichever comes first. This combination of full backups and continuous WAL archiving enables point-in-time recovery to any moment within the retention window.
Retention period
Backups are retained for 7 days, providing a sufficient window for recovery in case of data loss or corruption. If you need a longer backup retention period, contact support.
Storage and durability
Backups are replicated across multiple servers using erasure coding, ensuring they remain accessible even if some storage servers become unavailable. Backup storage is isolated at the bucket level—each Managed Postgres instance has its own dedicated storage bucket, and credentials are scoped to only allow access to that instance's backups.
Point-in-time recovery
Point-in-time recovery allows you to restore your database to any specific moment within the backup retention window. This is useful for recovering from accidental data deletion, corruption, or other issues where you need to roll back to a known good state.
To perform a point-in-time recovery:
- Navigate to the Backups view of your Managed Postgres instance.
- In the Point in time recovery section, select the target date and time (in UTC) you want to restore to.
- Click Restore to point in time.
The restore operation creates a new Managed Postgres instance with the state of your database as it was at the selected point in time. Your original instance remains unchanged, allowing you to verify the restored data before making any decisions about which instance to keep.
