Know the Constraints & Boundaries
64 TiB maximum storage per instance
Maximum 15 read replicas per source
35 days maximum retention period
63 characters maximum length
128 TiB maximum cluster volume size
15 Aurora replicas per cluster
5 secondary regions maximum
72 hours maximum backtrack window
Smallest instance type
Memory optimized
Largest instance
6.1 TiB per node maximum
500 nodes per cluster maximum
500 shards per cluster
65,000 concurrent connections
4.7 TiB per node maximum
40 nodes per cluster maximum
1 MB maximum per object
250 bytes maximum key length
400 KB maximum
255 characters max
1 MB maximum
25 items maximum
Not all services in all regions
Limited instance types in some regions
Data transfer costs apply
30 minutes minimum duration
Cannot overlap with maintenance
Time depends on data size
Problem: Encryption must be enabled at creation
Workaround: Create encrypted snapshot → Restore to new encrypted instance
Problem: Too many concurrent connections
Workaround: Use RDS Proxy or connection pooling
Problem: Database becomes read-only when storage full
Workaround: Enable storage autoscaling or monitor storage metrics
40 instances per region (adjustable)
50 per region (adjustable)
50 per region
Set up CloudWatch alarms
Right-size instances and storage
Use Service Quotas console