Database services

Run a database yourself on EC2 or let AWS manage it. Know which managed service fits relational, key-value, document, graph, and in-memory needs, and which tools migrate databases.

Self-managed vs. managed

• Database on EC2: Full control of the OS and database software (for example, an engine or version AWS doesn't offer as managed), but you do patching, backups, and high availability.
• AWS managed database: AWS handles provisioning, patching, backups, and replication, so you focus on schema and queries.

Relational

• Amazon RDS: Managed relational databases (MySQL, PostgreSQL, MariaDB, Oracle, SQL Server, Db2). Multi-AZ deployments for high availability; read replicas for read scaling.
• Amazon Aurora: AWS-built relational database compatible with MySQL and PostgreSQL, with high performance and storage replicated across AZs.

NoSQL and specialized

• Amazon DynamoDB: Serverless key-value and document NoSQL database with single-digit millisecond performance at any scale.
• Amazon DocumentDB: Managed document database with MongoDB compatibility.
• Amazon Neptune: Managed graph database for highly connected data (social networks, fraud detection, recommendations).

In-memory

• Amazon ElastiCache: Managed in-memory caching (Valkey, Redis OSS, Memcached) for microsecond reads, often placed in front of a database.

Database migration

• AWS DMS: Migrates databases to AWS with the source staying online; supports same-engine and different-engine migrations and continuous replication.
• AWS SCT: Converts schema and code objects when moving between different engines (for example, Oracle to Aurora PostgreSQL).