AWS SAA-C03 Study Guide: Chapter 10 – Cloud Architecture and AWS Framework

AWS Solutions Architect Associate (SAA-C03)

Chapter 10: Cloud Architecture & Well-Architected Framework

INFOGRAPHIC: Architecture at a Glance

1. The 6 Pillars of Well-Architected Framework

⚙️
Operational Excellence

🔒
Security

🏗️
Reliability

⚡
Performance Efficiency

💰
Cost Optimization

🌱
Sustainability

2. The 12-Factor App & AWS Mapping

1. CodebaseOne repo, many deploysAWS CodeCommit

2. DependenciesExplicitly isolateAWS CodeArtifact

3. ConfigStore in environmentSecrets Manager / Parameter Store

4. Backing ServicesTreat as attached resourcesRDS / SQS / DynamoDB

5. Build, Release, RunStrict separationCodeBuild / CodePipeline

6. ProcessesStateless executionLambda / Fargate

7. Port BindingExport services via portElastic Load Balancing

8. ConcurrencyScale out via process modelAuto Scaling

9. DisposabilityFast startup/graceful shutdownLambda

10. Dev/Prod ParityKeep environments similarCloudFormation / CDK

11. LogsTreat logs as event streamsCloudWatch Logs

12. Admin ProcessesRun as one-off tasksStep Functions

3. Popular Architecture Patterns

Tiered (N-Tier)

Separation of Presentation, App, and Data layers. Classic and robust.

Microservices

Decoupled small services. Scalable and independent deployment.

Event-Driven (EDA)

Reactive systems using Producers, Consumers, and Brokers (SNS/SQS).

Serverless

No server management. FaaS (Lambda) and BaaS (Amplify/Cognito).

Study Guide: Core Concepts

Exam Tip: Understand the difference between High Availability (staying operational) and Fault Tolerance (operating correctly even during a partial failure).

Key Resilience Metrics

RTO (Recovery Time Objective)

The “Downtime” clock. How quickly must you restore service?

RPO (Recovery Point Objective)

The “Data Loss” clock. How much data (in time) can you afford to lose?

Architectural Considerations

Elasticity: Automatically adjusting resources (Auto Scaling).
Decoupling: Removing dependencies using queues (SQS).
Loose Coupling: Components function independently; if one fails, others continue.
Design for Failure: Assume everything will fail eventually. Use Multi-AZ and Redundancy.

The Well-Architected Framework: Deep Dive

1. Operational Excellence

Focus: Running and monitoring systems. Key tool: CloudFormation (Infrastructure as Code).

2. Security

Focus: Protecting data and assets. Key tools: IAM, KMS (Encryption), GuardDuty.

3. Reliability

Focus: Ability to recover from disruptions. Key tools: Route 53 (DNS Failover), Multi-AZ RDS.

4. Performance Efficiency

Focus: Using resources efficiently. Key strategy: Right-sizing and choosing the correct instance types.

5. Cost Optimization

Focus: Avoiding unnecessary costs. Key tools: Cost Explorer, Trusted Advisor, Savings Plans.

6. Sustainability

Focus: Minimizing environmental impact. Key strategy: Maximizing utilization and reducing waste.

Data-Driven Architectures

Data Lake: Centralized repository for raw data (S3).
Data Warehouse: Structured data optimized for complex queries (Redshift).
Real-time Processing: Streaming data analysis (Kinesis).

### Key Highlights for the Exam: 1. **The 12-Factor App:** You will likely see questions regarding “Dev/Prod Parity” (solved by CloudFormation) and “Config” (solved by Secrets Manager). 2. **RTO vs. RPO:** Remember that a lower RTO/RPO usually increases costs because it requires faster failover and more frequent backups. 3. **WAF Pillars:** If a question asks about “Right-sizing,” the answer is usually **Cost Optimization** or **Performance Efficiency**. If it asks about “Infrastructure as Code,” think **Operational Excellence**. 4. **Serverless Limits:** Remember that AWS Lambda has a **15-minute** execution limit—it is not suitable for long-running batch jobs (use ECS or EC2 instead).