SAA
Multi-AZ High Availability Design for the SAA Exam
A beginner-friendly summary of Multi-AZ, redundancy, and single points of failure — concepts commonly tested in SAA-C03.
The Solutions Architect Associate exam often tests high availability design. Multi-AZ configurations, redundancy, and eliminating single points of failure are core topics you need to understand.
Summary
To improve availability, avoid designs that depend on a single AZ, a single instance, or a single component. In AWS, spanning multiple AZs is the standard approach.
What Is a Single Point of Failure?
A single point of failure (SPOF) is any component that brings down the whole system if it fails. Examples include a single EC2 instance, a database in only one AZ, or a NAT Gateway with no redundancy.
The SAA exam often asks questions like "What is the weak point of this design?" or "How can you improve its availability?"
What Is Multi-AZ?
Multi-AZ means using more than one Availability Zone. If one AZ fails, processing can continue in another AZ.
RDS Multi-AZ is a common pattern for making a database highly available. This is different from a Read Replica, which is designed to scale read performance.
ALB and Auto Scaling
For web applications, ALB distributes traffic across EC2 instances in multiple AZs, and Auto Scaling adjusts the number of instances. This makes the system resilient to both a single EC2 failure and an AZ-level failure.
High Availability for Serverless Services
Managed services such as Lambda, S3, DynamoDB, and API Gateway are designed by AWS with high availability built in. However, availability can still drop if you misconfigure settings or introduce a fragile dependency.
For example, if you place a Lambda function inside a VPC and route its traffic through a NAT Gateway, you need to think about NAT Gateway redundancy as well.
Key Points for the Exam
- A single-AZ design is vulnerable to failures
- Multi-AZ is used to improve availability
- Read Replicas are mainly used to scale read performance and distribute read traffic
- ALB + Auto Scaling provides redundancy at the web layer
- Even with managed services, check your dependencies
Related Terms
Further Reading
What to learn next
Combine blog posts with the AWS glossary, service comparisons, practice questions, and architecture gallery to connect exam knowledge with implementation.