How to Design and Implement Microservices: Claude Code 활용 가이드

Where Claude Code Shines in Microservices Development

Microservices architecture requires careful consideration of inter-service communication, data consistency, and shared library management. Claude Code can make consistent changes across multiple services while maintaining a bird’s-eye view of the entire project.

Service Decomposition Design

> I want to split an e-commerce site into microservices.
> Using domain-driven design bounded contexts as a reference,
> propose a service decomposition plan. Design the APIs for each service.

Claude Code proposes a service structure like the following:

• user-service: User management and authentication
• product-service: Product catalog management
• order-service: Order processing
• payment-service: Payment processing
• notification-service: Notification delivery

Inter-Service Communication

Synchronous Communication via REST API

// order-service/src/clients/product-client.ts
import axios from 'axios';

const PRODUCT_SERVICE_URL = process.env.PRODUCT_SERVICE_URL || 'http://product-service:3001';

export class ProductClient {
  async getProduct(productId: string) {
    const { data } = await axios.get(
      `${PRODUCT_SERVICE_URL}/api/products/${productId}`,
      { timeout: 5000 }
    );
    return data;
  }

  async checkStock(productId: string, quantity: number): Promise<boolean> {
    const { data } = await axios.post(
      `${PRODUCT_SERVICE_URL}/api/products/${productId}/check-stock`,
      { quantity },
      { timeout: 5000 }
    );
    return data.available;
  }
}

Asynchronous Communication via Event-Driven Architecture

// order-service/src/events/publisher.ts
import { SNSClient, PublishCommand } from '@aws-sdk/client-sns';

const sns = new SNSClient({ region: 'us-east-1' });

export async function publishOrderCreated(order: Order) {
  await sns.send(new PublishCommand({
    TopicArn: process.env.ORDER_EVENTS_TOPIC_ARN,
    Message: JSON.stringify({
      type: 'ORDER_CREATED',
      data: {
        orderId: order.id,
        userId: order.userId,
        items: order.items,
        totalAmount: order.totalAmount,
      },
      timestamp: new Date().toISOString(),
    }),
    MessageAttributes: {
      eventType: { DataType: 'String', StringValue: 'ORDER_CREATED' },
    },
  }));
}

// notification-service/src/handlers/order-handler.ts
import { SQSEvent } from 'aws-lambda';

export async function handleOrderEvent(event: SQSEvent) {
  for (const record of event.Records) {
    const message = JSON.parse(record.body);
    const orderEvent = JSON.parse(message.Message);

    if (orderEvent.type === 'ORDER_CREATED') {
      await sendOrderConfirmationEmail(orderEvent.data);
    }
  }
}

Docker Compose Configuration

> Set up docker-compose so all services can be started together.
> Include hot reload support, shared network, and health checks.

version: '3.8'

services:
  user-service:
    build: ./services/user-service
    ports: ['3001:3000']
    environment:
      - DATABASE_URL=postgresql://postgres:pass@user-db:5432/users
    depends_on:
      user-db:
        condition: service_healthy
    volumes:
      - ./services/user-service/src:/app/src
    healthcheck:
      test: ['CMD', 'curl', '-f', 'http://localhost:3000/health']
      interval: 10s
      retries: 3

  product-service:
    build: ./services/product-service
    ports: ['3002:3000']
    environment:
      - DATABASE_URL=postgresql://postgres:pass@product-db:5432/products
    depends_on:
      product-db:
        condition: service_healthy
    volumes:
      - ./services/product-service/src:/app/src

  order-service:
    build: ./services/order-service
    ports: ['3003:3000']
    environment:
      - DATABASE_URL=postgresql://postgres:pass@order-db:5432/orders
      - PRODUCT_SERVICE_URL=http://product-service:3000
      - USER_SERVICE_URL=http://user-service:3000
    depends_on:
      order-db:
        condition: service_healthy

  user-db:
    image: postgres:16-alpine
    environment:
      POSTGRES_DB: users
      POSTGRES_PASSWORD: pass
    healthcheck:
      test: ['CMD-SHELL', 'pg_isready -U postgres']
      interval: 5s
      retries: 5

  product-db:
    image: postgres:16-alpine
    environment:
      POSTGRES_DB: products
      POSTGRES_PASSWORD: pass
    healthcheck:
      test: ['CMD-SHELL', 'pg_isready -U postgres']
      interval: 5s
      retries: 5

  order-db:
    image: postgres:16-alpine
    environment:
      POSTGRES_DB: orders
      POSTGRES_PASSWORD: pass
    healthcheck:
      test: ['CMD-SHELL', 'pg_isready -U postgres']
      interval: 5s
      retries: 5

Shared Library Management

> Create shared type definitions and utilities in packages/shared.
> Use npm workspaces for management.

// packages/shared/src/types/events.ts
export interface DomainEvent<T = unknown> {
  type: string;
  data: T;
  timestamp: string;
  correlationId: string;
}

export interface OrderCreatedEvent extends DomainEvent<{
  orderId: string;
  userId: string;
  items: Array<{ productId: string; quantity: number }>;
  totalAmount: number;
}> {
  type: 'ORDER_CREATED';
}

Tips for Improving Development Efficiency

For microservices development, documenting your service structure and API conventions in CLAUDE.md is effective. For details, see the Complete CLAUDE.md Guide. Also leverage Refactoring Automation to maintain code quality. Setting up hooks per service development flow is also useful.

정리

With Claude Code, you can efficiently build everything from microservices design to implementation, Docker configuration, and CI/CD. It’s especially powerful for cross-service tasks like sharing type definitions and maintaining event schema consistency.

For more details, see the official Anthropic documentation.