A snapshot preserves data from an earlier state so it can be restored after accidental deletion or corruption. Durable storage is not a substitute: replicated data can still reflect a mistaken deletion. In AWS Backup, a snapshot is called a recovery point. section 9.6 section 12.10
Choose how snapshots are created and kept:
- For EFS, use AWS Backup: a plan schedules backups, a vault groups recovery points, and a selection identifies the filesystem. An IAM role grants AWS Backup access. section 9.6
- For RDS, configure automated snapshots and their retention period, or trigger a manual snapshot before a risky change such as a WordPress upgrade or schema migration. Automated snapshots expire after retention; manual snapshots remain until you delete them. section 10.3 section 10.3.2
- For DynamoDB tables, use on-demand backups to preserve a working state before data is deleted or corrupted. section 12.10
Create an RDS snapshot with the AWS CLI after obtaining the database instance identifier. Snapshot creation takes a few minutes; check its state before trying to restore it. section 10.3.2
Restoration differs by service. AWS Backup can restore an entire EFS filesystem or selected folders, into the original or a new filesystem. Restoring an RDS snapshot creates a new database instance; it cannot overwrite the existing one. An application can then be pointed at the restored database’s endpoint. RDS automated backups also support restoration to a specified time within the retention period using database change logs. section 9.6 section 10.3.3
A copied RDS snapshot can support recovery or relocation in another AWS region. Check privacy and compliance rules before moving real data. Snapshots use billable storage, so delete manual and copied snapshots when they are no longer needed. section 10.3.4 section 10.3.5











