Amazon Aurora DSQL
Serverless distributed SQL
A serverless, distributed, PostgreSQL-compatible relational database; it automatically scales reads/writes/storage, has no servers to manage, and supports multi-Region active-active clusters with strong consistency.
- Cost$$ Medium
- Difficulty●●●●●
- Costs money when idle?A little
- Serverless
Think of it asLike a shared ledger kept in many cities: whatever is written in one place is immediately correct everywhere.
When to use it
✓ Use it when
Transactional applications that need SQL, very high availability and multi-Region writes without managing instances.
✗ Avoid it when
You need full PostgreSQL features and extensions → Aurora PostgreSQL / RDS; simple key-value → DynamoDB.
Common uses
- Multi-Region financial and payment applications
- Auto-scaling SaaS backend
- Microservices needing SQL without managing servers
What does it cost?
- Pricing model: Charged by DPUs (distributed processing units) consumed for reads/writes/compute and GB-month of storage.
⚠️ Cost traps
Multi-Region clusters are charged for storage in each Region plus cross-Region data transfer.
