Amazon RDS
Amazon Relational Database Service
Overview
Amazon RDS makes it easy to set up, operate, and scale relational databases in the cloud. Supports MySQL, PostgreSQL, MariaDB, Oracle, and SQL Server — AWS handles OS and database installation, patching, backups, and failover. You focus on schema design and queries rather than database administration. High availability and backup are enabled through configuration.
Key Concepts
Multi-AZ
With Multi-AZ enabled, a synchronously replicated standby is created in a different AZ. If the primary fails, automatic failover to the standby occurs and applications reconnect using the same endpoint. This is an availability (fault tolerance) feature — it does not improve read performance.
Read Replicas
Read replicas use asynchronous replication to create read-only copies that offload read-heavy queries and improve read scalability. Can be created in the same or a different region for cross-region DR and global read latency. Multi-AZ (availability) vs. Read Replica (scalability) is the most-tested distinction in RDS questions.
Backup and Recovery
Automated backups with transaction logs enable point-in-time recovery (PITR) to any second within the retention period. Manual snapshots persist until explicitly deleted and can be copied to another region or used to create a new instance.
Storage and Scaling
Choose General Purpose SSD or Provisioned IOPS storage; storage auto-expansion is configurable. Compute scales vertically by changing instance class. Write scaling has limits, so consider Aurora or a caching layer (for read load) when needed.
Security and Encryption
RDS runs inside a VPC; security groups control access. Data at rest uses KMS encryption; data in transit uses SSL/TLS. Credentials are safely managed and auto-rotated via Secrets Manager. Always place RDS in private subnets — never expose it directly to the internet.
Typical Use Cases
- Run a highly available relational database backend for web applications
- Use Multi-AZ to minimize downtime during planned maintenance or AZ failures
- Distribute read-heavy reporting and analytics queries to Read Replicas
- Use automated backups and PITR to recover from accidental changes or data corruption
- Migrate on-premises databases to managed cloud infrastructure to reduce operational burden
Exam Relevance
RDS is heavily tested in SAA, SOA, and DVA. The single most important distinction: Multi-AZ = availability (automatic failover) vs. Read Replica = read scalability. Misidentifying the goal (HA vs. performance) leads directly to wrong answers. Classic chains: 'high read load → Read Replica,' 'write scaling + global → Aurora,' 'millisecond caching → ElastiCache.' SOA covers backup retention and PITR. SAP covers cross-region replicas for DR. DVA covers Secrets Manager for credential management.
Related Services
References
Last updated: 2026-06-24
This page is created from AWS official documentation and reviewed/edited by the operator. See our editorial & quality policy for details.