3.1.4. Amazon RDS, Aurora, and Relational Options
💡 First Principle: Relational databases excel where data relationships and ACID transactions are paramount — but they reach scalability limits that purpose-built analytics stores don't have. When the exam describes transactional workloads (order processing, user authentication, inventory management), relational databases are the answer. When it describes analytical workloads on the same data, the answer usually involves a pipeline from the relational database to an analytics store.
Amazon RDS manages relational database engines: MySQL, PostgreSQL, MariaDB, Oracle, SQL Server. RDS handles provisioning, patching, backups, and failover. Key features for the exam: Multi-AZ deployment (synchronous standby for high availability), read replicas (asynchronous copies for read scaling), and automated backups with point-in-time recovery.
Amazon Aurora is AWS's cloud-native relational database, compatible with MySQL and PostgreSQL. Aurora's architecture separates compute from storage, with a distributed storage layer that replicates data 6 ways across 3 AZs. Key advantages: up to 5x faster than MySQL, up to 3x faster than PostgreSQL, automatic storage scaling up to 128 TB, and up to 15 read replicas with sub-10ms replication lag.
Aurora Serverless v2 scales compute capacity automatically between a minimum and maximum ACU (Aurora Capacity Unit). Ideal for variable workloads, dev/test, and applications with intermittent usage. The exam v1.1 update added Aurora to the in-scope service list, so expect questions on when to choose Aurora over standard RDS.
RDS Proxy pools and shares database connections, reducing the connection overhead that Lambda functions create (each invocation opens a new connection). If a question describes Lambda-to-RDS integration with connection exhaustion issues, RDS Proxy is the answer.
⚠️ Exam Trap: Aurora and RDS are both relational, but they're different services with different architectures. If a question mentions "cloud-native relational database" or "MySQL/PostgreSQL compatible with automatic storage scaling," it's Aurora. If it mentions a specific engine not supported by Aurora (Oracle, SQL Server), it's RDS.
Reflection Question: A microservices application uses Lambda functions that each open a connection to an Aurora PostgreSQL database. During peak load, the database hits its connection limit. What AWS service resolves this without changing application code?
