RDS and Aurora
Amazon Relational Database Service (RDS) runs conventional relational engines as a managed service: AWS owns the host, the patching, the backups and the failover, and you own the schema and the queries. Amazon Aurora is AWS’s own MySQL- and PostgreSQL-compatible engine, built on a distributed storage layer, and is offered through the same console and API.
These pages cover both.
The services
Section titled “The services”- Amazon RDS — supported engines, Multi-AZ, read replicas, backups, security, and the cases where RDS is the wrong answer.
- Amazon Aurora — the storage architecture, endpoints, replicas, global databases and how Aurora differs from RDS.
- Aurora Serverless — capacity units, automatic scaling, and scaling to zero.
Operating them
Section titled “Operating them”- Scaling relational databases — the four ways to add capacity, and what each costs.
- Disaster recovery for RDS — strategies from backup-and-restore to multi-Region, with their recovery objectives.
- Multi-AZ and RDS Proxy — two independent features that are frequently confused.
- Blue/green deployments — upgrading an engine or changing a schema with a short, reversible switchover.
- Sharing a database with an external auditor — the snapshot and KMS mechanics of giving another account a copy.
If you are deciding whether a relational database is the right choice at all, start with OLTP vs OLAP and the service comparison.