RDS

Think database + EC2.

Can use SQL (like MySQL, PostgresSQL, Oracle, Microsoft SQL server).

Comes with automatic patching, backups, redundency, failover, and disaster recovery.

Aurora

Amazon's own type of database. It's not MySQL or Postgres, but it quacks like it.

Data is replicated across AZs(?) with 6 copies. 15 read replicas. Continus backups to S3. Point-in-time recovery.

DynamoDB

Key-value DB. Think back to the mIRC tables, these are exactly the same.

Also serverless.

Redshift

Data warehousing/big data. Not serverless.

Database Migration Service (DMS)

Move database (sql or not) to AWS.

Usecases

  • Development and test migration - You can use DMS to make a copy of production data to dev enviorment to test against without affecting prod.
  • Consolidation - merge several databases into one.
  • Continuous replication - send ongoing copies to other targets.

Homogenous

Source and target are the same type.

  • MySQL, Microsoft SQL to EC2 or RDS

Hetrogenous

Source and target are of different types.

  • Need to be converted using the conversion tool. Will change the schema and code to match target.
  • Use DMS to target.

DocumentDB

Think content management system.

Neptune

Graphing database. Who is connected to who kind. For social networking and recomendations and fraud detection.

QLDB

Think financial actions tracking. Any entry can never be removed.

Managed Blockchain

For blockchains.

ElastiCache

In memory database.

DAX

Caching for DynamoDB.