Skip to main content

Integration Styles

Integration Styles covers the architecture layer defining how services communicate and cooperate with each other and external systems.

✅ Problems Addressed by This Category​

  • How to choose communication methods between services.
  • Latency and scale problems of synchronous APIs.
  • Complexity of event-driven architecture.
  • Handling distributed transactions with Saga.
  • Contract management of API boundaries.

✅ Representative Styles of Integration Styles​

● REST / gRPC / GraphQL​

  • Basic methods of synchronous communication.
  • REST: Most widely spread, for CRUD.
  • gRPC: Fast, type-safe, for Microservices.
  • GraphQL: Frontend optimization (Good compatibility with BFF).

● Event-Driven Architecture (EDA)​

  • Loose coupling with asynchronous events.
  • Suitable for high scale / high availability configuration.
  • However, complexity of observability and consistency increases.

● Saga Pattern (Orchestration / Choreography)​

  • Control of distributed transactions.
  • Essential for "consistency across multiple services" like payment and reservation.

● API Gateway / BFF​

  • Integration and optimization of APIs for clients.
  • Aggregates authentication, routing, and rate limiting.

● Service Mesh​

  • Separates communication from app and controls it at network layer.
  • Automates mTLS, retry, Circuit Breaker, tracing, etc.

● Outbox Pattern (Transactional Outbox)​

  • Buffers events within DB transaction.
  • Resolves inconsistency between app update and event delivery.
  • Most practical consistency assurance method in Microservices.

● Change Data Capture (CDC)​

  • Directly streams DB change logs.
  • Generates events without going through the app.
  • Suitable for high throughput data replication and pipelines.

✅ How to Read This Category​

  1. Distinction between Synchronous / Asynchronous.
  2. Trade-off between "Consistency" and "Latency".
  3. Role of layers absorbing complexity like BFF and Mesh.

✅ Relationship with Other Layers​