Advanced Aurora Concepts for the AWS Exam

1. Replica Auto Scaling

Replica Auto Scaling automatically adjusts the number of Amazon Aurora replicas based on workload demand.

How It Works

• A client connects to an Aurora cluster with one writer instance (using the writer endpoint) and multiple reader instances (using the reader endpoint).
• If read requests increase and CPU usage rises, Aurora adds replicas automatically.
• The reader endpoint expands to include these new replicas, distributing read traffic efficiently and reducing CPU load.

2. Custom Endpoints

Custom endpoints allow you to define subsets of Aurora replicas for specific workloads.

Use Case

• Some replicas may have larger instance sizes (e.g., db.r5.2xlarge vs. db.r3.large).
• A custom endpoint can be assigned to the more powerful replicas for running analytical queries.
• Once a custom endpoint is in use, the general reader endpoint is no longer needed.
• Multiple custom endpoints can be defined for different workloads.

3. Aurora Serverless

Aurora Serverless provides automated database instantiation and scaling based on actual usage.

Key Benefits

• Ideal for infrequent, intermittent, or unpredictable workloads.
• Eliminates capacity planning—instances scale automatically based on demand.
• Pricing is based on per-second usage, making it a cost-effective solution.