Why it matters
Automated backups and manual snapshots for RDS and Aurora are billed per GB-month beyond what is covered by free backup storage, and can quietly grow into a significant share of your bill. Old snapshots taken for one-off changes, long retention periods, and cross-region copies all add recurring cost without improving day-to-day resilience.
Recurring backup hygiene checklist
Every quarter (or at least twice a year), run a quick review:
- Retention sanity check – Confirm automated backup retention for each environment matches current RPO/RTO and compliance needs (no “maxed out just in case” settings).
- Snapshot cleanup – List manual snapshots (including cross-region copies) and delete those no longer needed for recovery, investigations, or audits.
- Big-spender review – In Cost Explorer, identify the largest RDS/Aurora backup line items and verify each has a clear owner and justification.
How backups are charged
-
Automated backups
- RDS includes free backup storage up to the size of your provisioned database storage; beyond that, backup data is charged per GB-month.
- Retention is configured per-instance/cluster (for example, 7–35 days). Longer retention increases backup storage consumption.1
-
Manual snapshots
- Charged per GB-month for as long as the snapshot exists, regardless of the instance’s current state.
- Cross-region or cross-account snapshot copies incur additional storage and data transfer costs.1
Aurora follows similar patterns, but because storage automatically grows, backup and snapshot costs can track that growth unless you prune old restore points.2
How to identify in Cost Explorer
RDS
- Filter by Service → RDS and Usage Type → RDS ChargedBackup
- Group by → Usage Type
RDS backup costs filtered by ChargedBackup usage type in Cost Explorer
Aurora
- Filter by Service → Relational Database Service (RDS) , Usage Type → look for Aurora BackupUsage and select all
- Group by → Usage Type
Aurora backup costs filtered by BackupUsage usage types in Cost Explorer
Quick Wins
-
Align retention with RPO/RTO
- Set automated backup retention to what you actually need for recovery and compliance (for example, 7–14 days for many application workloads, longer for regulatory archives).
- Use point-in-time recovery plus a smaller number of longer-lived manual snapshots instead of “forever” retention on every instance.
-
Clean up old manual snapshots
- Regularly list snapshots and delete those that were created for past migrations, upgrades, or experiments and no longer serve a purpose.
- Pay particular attention to large production databases and cross-region copies, which can be some of the most expensive snapshots to keep.
-
Use tags and automation
- Tag snapshots with owner, environment, purpose, and expiry date so you can safely identify which ones can be removed.
- Use scheduled automation (for example, AWS Backup policies or Lambda/Step Functions jobs) to enforce snapshot creation and deletion rules rather than relying on ad hoc manual cleanup.
-
Be deliberate with cross-region backups
- Only enable cross-region snapshot copy where you have a clear requirement for regional disaster recovery or data residency.
- Choose target regions carefully based on latency and storage pricing; consider whether fewer clusters with well-tested restore procedures can meet your objectives.
-
Review backup coverage after major changes
- When you add new RDS/Aurora clusters, change instance classes, or refactor environments, confirm that backup policies and retention still match expectations and that you are not backing up ephemeral or short-lived test data for long periods.
Related strategies
- Optimize storage – Control primary database storage costs alongside backup storage for a comprehensive storage cost strategy.