AWS managed databases — RDS (PostgreSQL/MySQL), DynamoDB, Aurora, ElastiCache (Redis).
Stub — full skill pending. Core patterns documented below.
| Service | Type | Use case |
|---|---|---|
| RDS PostgreSQL | Relational | General purpose; preferred over MySQL for new projects |
| RDS MySQL | Relational | Required for Frappe/ERPNext |
| Aurora Serverless v2 | Relational | Auto-scales; good for variable load; PostgreSQL-compatible |
| DynamoDB | NoSQL key-value | High-throughput event storage, session store, leaderboards |
| ElastiCache Redis | In-memory | Session cache, queue, rate limiting, pub/sub |
# Create RDS PostgreSQL instance
aws rds create-db-instance \
--db-instance-identifier mydb \
--db-instance-class db.t3.micro \
--engine postgres \
--engine-version 16.2 \
--master-username admin \
--master-user-password $(aws secretsmanager get-secret-value --secret-id db-password --query SecretString --output text) \
--allocated-storage 20 \
--vpc-security-group-ids sg-xxxxxxxx \
--db-subnet-group-name my-subnet-group \
--backup-retention-period 7 \
--no-publicly-accessible \
--region af-south-1Use AWS Secrets Manager to store DB credentials. Retrieve in application code — never hardcode.
For Lambda with RDS: consider RDS Proxy to pool connections (Lambda can exhaust DB connections under load — RDS Proxy prevents this).