Databases (RDS)
Audit your RDS footprint to eliminate costs from abandoned or underutilized databases.
Permissions Required:
rds:DescribeDBInstances, rds:DescribeDBSnapshots, cloudwatch:GetMetricStatistics.RDS Instances
Stopped Instances
AWS Doctor identifies RDS instances that are in a stopped state.
- Reason: While compute costs are paused, you are still billed for the allocated storage and any associated IOPS or multi-AZ configuration.
- Action: If the instance is no longer needed, consider deleting it after taking a final snapshot.
Idle RDS Instances
Finds instances that are in an available state but have had zero database connections for the last 7 days.
- Reason: These instances are fully operational and billing for compute and storage but are not being utilized by any application.
- Action: Stop the instance if it’s only needed occasionally, or terminate it if it’s no longer used.
You can customize the 7-day idle threshold using the
--rds-idle-days flag.RDS Snapshots
Old Manual Snapshots
Identifies manual RDS snapshots that are older than 30 days.
- Reason: Unlike automated snapshots which are deleted according to a retention policy, manual snapshots persist indefinitely until manually deleted, incurring ongoing storage costs.
- Action: Review old snapshots and delete those that are no longer required for compliance or disaster recovery.
You can customize the 30-day age threshold using the
--rds-snapshot-days flag.