<< back to Guides

🛡️ Guide: API Gateway — A Systems Design Deep Dive

An API Gateway is a centralized entry point that manages, routes, secures, and scales API requests in a microservices architecture.

It decouples clients from backend services, enabling simplified API access, load management, security, and observability.

🧠 1. What Is an API Gateway?

An API Gateway acts as a reverse proxy between clients and backend services.

It:

Client → API Gateway → Auth Service
                    → User Service
                    → Payment Service

🎯 2. Why Use an API Gateway?

Purpose Benefit
Decoupling Clients are not tied to internal service structure
Security Central place to enforce authentication, TLS
Rate Limiting Prevents overload, abuse
Routing & Load Balancing Directs traffic efficiently
Protocol Translation Converts REST ↔ gRPC ↔ WebSockets
Aggregation Combines responses from multiple services
Monitoring Unified logging and metrics

🧱 3. Core Components

Component Role
Request Router Routes to appropriate backend
Authentication Layer Verifies tokens, keys
Rate Limiter Prevents overuse per client or route
Caching Layer Stores frequent responses
Transformation Layer Request/response shaping (e.g., JSON → XML)
Logging & Analytics Centralized observability

🧰 4. API Gateway vs Load Balancer

Feature API Gateway Load Balancer
Layer Application layer (L7) Transport layer (L4/L7)
Aware of APIs Yes No
Auth, Throttling Built-in Not available (unless L7)
Protocol Translation Yes (REST ↔ gRPC, etc.) No
Routing Logic Fine-grained (method/path/service) IP/port-based

🧪 5. API Gateway Patterns

🛠️ 1. Simple Gateway

Routes requests directly based on path or method:

GET /users → User Service  
POST /payments → Payment Service

🔐 2. Gateway with Auth & Rate Limiting

// Request Flow:
→ Verify JWT Token  
→ Check quota in Redis  
→ Forward to service  
→ Log + return response

📦 3. Backend for Frontend (BFF)

Custom gateway per frontend (mobile, web, admin):

📊 4. Aggregator Gateway

Combines multiple backend calls into one response:

GET /profile → calls:
  /user/123
  /user/123/orders
  /user/123/preferences
→ returns combined JSON

⚙️ 6. API Gateway Architecture Diagram

               ┌───────────────┐
Client ───────▶│ API Gateway   │────▶ Auth Service
               │               │────▶ User Service
               │               │────▶ Order Service
               └───────────────┘
                      │
                Logging & Metrics

🧱 7. Real-World Gateways

Tool/Platform Features
Kong Open source, plugin-based, Lua core
NGINX Lightweight reverse proxy, Lua scripting available
Envoy L7 proxy with native gRPC support
AWS API Gateway Fully managed, integrates with Lambda, IAM, etc.
Istio Ingress Gateway Part of service mesh, tightly integrated with Envoy
Zuul (Netflix) Java-based, Spring Cloud integrated

🧠 8. Common Design Considerations

Concern Solution / Notes
Latency Add caching, keep transformations minimal
Single point of failure Deploy gateway replicas + load balancer
Scaling Use stateless design, autoscale pods
Auth overhead Use opaque token + caching
Rate limiting Global counters in Redis or in-memory

🧪 9. Example: Kong Gateway Rate Limiting Plugin (Declarative Config)

plugins:
- name: rate-limiting
  config:
    minute: 100
    policy: local

✅ 10. Summary

Feature Why It Matters
Entry point One place to manage all API traffic
Central policies Auth, rate limiting, monitoring
Decoupled services Clients stay unaware of backend complexity
Scalable Stateless, horizontally scalable
Flexible Works across REST, GraphQL, gRPC, WebSockets

📚 Further Reading

<< back to Guides