Architecture patterns are where individual decisions become complete systems.
While pillars explain how to make specific decisions and topics organize what decisions you’ll face, patterns show how everything fits together in production-ready architectures.
These aren’t theoretical patterns. They’re the recurring enterprise architectures that AWS Solutions Architects design repeatedly—and that SAP-C02 tests repeatedly.
The Serverless & Microservices pattern is the cornerstone of modern, cloud-native application design. In the SAP-C02 exam, the focus shifts from managing servers to orchestrating distributed systems that are cost-effective, highly available, and capable of handling massive spikes in traffic without manual intervention.
🏗️ Core Architectural Patterns #
SAP-C02 evaluates your ability to build resilient microservices by applying the principles of the Well-Architected Framework.
1. The “Serverless Trio” (API + Compute + Data) #
- API Gateway: Acts as the “front door,” handling authentication, throttling, and request transformation.
- AWS Lambda: Provides the compute logic. Key exam focus: Concurrency limits, Cold starts, and Dead Letter Queues (DLQ).
- Amazon DynamoDB: The preferred NoSQL database for serverless, offering seamless scaling and event-driven triggers (DynamoDB Streams).
2. Asynchronous Orchestration vs. Choreography #
- AWS Step Functions (Orchestration): Best for complex workflows with state management, retry logic, and human-in-the-loop steps.
- Amazon EventBridge (Choreography): The preferred “Serverless Event Bus” for decoupled, event-driven communication between multiple microservices.
- Amazon SQS/SNS: Traditional pub/sub and queuing to buffer traffic and ensure message durability.
📚 Key Decision Pillars (Deep Dives) #
Mastering these pillars is essential for designing “Zero Administration” systems:
- Serverless Scaling: Lambda & API GW: Deep dive into reserved vs. provisioned concurrency.
- Event-Driven Apps via EventBridge: Designing cross-account event buses.
- SQS, SNS & Kinesis Decoupling Logic: When to choose polling (SQS) vs. streaming (Kinesis).
- ECS vs EKS Orchestration Trade-offs: Knowing when a scenario outgrows Lambda and requires containers.