Idle RDS Instances

Why This Matters

Idle RDS instances are among the most expensive forms of AWS waste. Unlike EC2 instances, RDS includes managed service costs, backups, and monitoring - meaning even small instances cost significantly more. Zero database connections indicates the instance serves no purpose.

Common causes:

• Development databases that are no longer actively used
• Testing instances that weren't cleaned up after projects
• Backup databases that were promoted but original wasn't decommissioned
• Legacy applications that were migrated but databases left running

How to Identify Idle RDS Instances

AWS Saver flags RDS instances as idle when they meet these criteria:

1. Zero database connections averaged over 24 hours
2. Instance status is 'available' (running and accessible)
3. Aged over 7 days (avoids flagging new deployments)
4. Monthly cost above $50 (meaningful savings threshold for RDS)

How to Fix Idle RDS Instances

Step 1: Find available RDS instances

aws rds describe-db-instances \
  --query 'DBInstances[?DBInstanceStatus==`available`].[DBInstanceIdentifier,DBInstanceClass,Engine,InstanceCreateTime]'

Step 2: Check database connections over 14 days

aws cloudwatch get-metric-statistics \
  --namespace AWS/RDS \
  --metric-name DatabaseConnections \
  --dimensions Name=DBInstanceIdentifier,Value=your-db-instance \
  --statistics Average \
  --start-time $(date -u -d '1 day ago' +%Y-%m-%dT%H:%M:%SZ) \
  --end-time $(date -u +%Y-%m-%dT%H:%M:%SZ) \
  --period 3600

Step 3: Create final snapshot (recommended)

aws rds create-db-snapshot \
  --db-instance-identifier your-db-instance \
  --db-snapshot-identifier manual-snapshot-before-stop-$(date +%Y%m%d)

Step 4: Stop database instance (reversible)

aws rds stop-db-instance --db-instance-identifier your-db-instance

Step 5: Delete if truly unused (after 30 days stopped)

aws rds delete-db-instance \
  --db-instance-identifier your-db-instance \
  --final-db-snapshot-identifier final-snapshot-$(date +%Y%m%d) \
  --skip-final-snapshot false

Prevention Tips

Use scheduled start/stop: Configure development databases to automatically stop during off-hours using Lambda functions.

Tag resources properly: Use consistent tagging for environment (dev/test/prod) to track database lifecycle and ownership.

Set up monitoring: Create CloudWatch alarms for connection counts to catch idle databases early.

Implement lifecycle policies: Establish clear retention and cleanup policies for non-production databases.

Automation Available

Skip the manual work. AWS Saver automatically monitors RDS instances using the same zero-connection and 24-hour analysis criteria.

✅ Connection monitoring - Tracks DatabaseConnections metric over 24-hour periods

✅ Conservative thresholds - Only flags instances aged over 7 days and above $50/month cost

✅ CPU analysis - Also detects oversized RDS instances with under 20% CPU utilization

✅ Cost impact calculation - Shows exact monthly waste per idle instance