Database Disaster Recovery

CiviForm's production database is backed up daily, with snapshot names prefixed with the value set by POSTGRES_RESTORE_SNAPSHOT_IDENTIFIER. By default they are retained for 7 days, which can be configured byPOSTGRES_BACKUP_RETENTION_DAYS.

(AWS) Restoring the database to a previous snapshot

In the event of catastrophic data loss or corruption, the production database can be restored to a previous snapshot. To do this:

1. Identify the DB snapshot to restore to in the AWS console by navigating to RDS > Snapshots.

2. Copy the snapshot identifier (not the name, the identifier begins with rds:)

3. In your civiform_config.sh file, set POSTGRES_RESTORE_SNAPSHOT_IDENTIFIER to the identifier.

4. Remove deletion protection from the database by navigating to RDS > Databases, selecting your production database and removing deletion protection.

5. Run bin/deploy. You should see in the terraform plan that the RDS database will be replaced.

Note that once the database is restored, the POSTGRES_RESTORE_SNAPSHOT_IDENTIFIER should continue to reference the same snapshot identifier. Changing the value of it will trigger another database replacement which may result in data loss.