docker-best-practices

name	docker-best-practices
description	Docker 多服务部署最佳实践 - 遵循生产级 Dockerfile 和 Docker Compose 架构原则。核心原则：Dockerfile 不随环境变化、compose 不包含真实 secret、所有环境差异通过 .env 管理、容器可销毁但数据不可、恢复能力优于自动化炫技。涵盖：Dockerfile 优化、多服务 compose 架构、健康检查、数据持久化、安全配置。

Docker-Best-Practices Skill

Docker 多服务部署最佳实践 - 生产级 Dockerfile 和 Docker Compose 架构指南。

When to Use This Skill

This skill should be triggered when:

• Designing multi-service Docker architectures for production deployment
• Writing or optimizing Dockerfiles for size, security, and performance
• Configuring Docker Compose for multi-container applications
• Implementing data persistence strategies with volumes and mounts
• Setting up environment-specific configurations using .env files
• Implementing container health checks and monitoring
• Securing Docker containers with proper user permissions and capabilities
• Optimizing container resource usage (CPU, memory, I/O limits)
• Deploying applications from development to production with Docker
• Troubleshooting Docker deployment issues related to networking, storage, or configuration

核心架构运行哲学

必须遵守的 5 条原则：

1. Dockerfile 不随环境变化

   • Dockerfile 只定义"构建什么"，不定义"运行时配置"
   • 所有环境差异（dev/staging/prod）通过 .env 文件注入
   • 同一个 Dockerfile 应该能在所有环境使用

2. Compose 不包含真实 secret

   • compose.yaml 只包含 secret 的占位符或引用
   • 真实密钥从 .env、secret manager 或挂载文件读取
   • 永远不要将 compose.yaml 包含的 secret 提交到版本控制

3. 所有环境差异 → .env

   • 环境变量优先级：Shell > --env-file > .env > compose defaults
   • 每个环境有独立的 .env.{environment} 文件
   • .env.example 提供模板，但不包含真实值

4. 容器可销毁，数据不可

   • 容器是 ephemeral（临时的），随时可以删除重建
   • 数据必须持久化到 volumes 或外部存储
   • 状态与容器分离，重建不影响数据

5. 恢复能力 > 自动化炫技

   • 优先考虑故障恢复和数据安全
   • 避免过度复杂的自动化脚本
   • 简单的备份/恢复流程胜过复杂的编排

Dockerfile 构建五原则

必须遵循的 5 条构建原则：

1. 为合成而构建，把不变的放在易变的前面

   # 不变的底层依赖（很少变化）
   FROM python:3.12-slim AS base
   WORKDIR /app
   
   # 依赖安装（偶尔变化）
   COPY requirements.txt .
   RUN pip install --no-cache-dir -r requirements.txt
   
   # 应用代码（经常变化）
   COPY . .
   

2. 分离构建和运行，让最终的产物绝对纯粹

   # 构建阶段：包含所有构建工具
   FROM node:20 AS builder
   WORKDIR /app
   COPY package*.json ./
   RUN npm ci
   COPY . .
   RUN npm run build
   
   # 运行阶段：只包含运行时必需
   FROM node:20-slim
   COPY --from=builder /app/dist ./dist
   COPY --from=builder /app/node_modules ./node_modules
   USER node
   CMD ["node", "dist/server.js"]
   

3. 最小权限运行，永远不要相信默认设置

   FROM alpine:3.19
   # 创建非 root 用户
   RUN addgroup -g 1000 appgroup && \
       adduser -D -u 1000 -G appgroup appuser
   # 只开放必需的端口
   EXPOSE 8080
   # 以非 root 用户运行
   USER appuser
   # 删除不必要的 capabilities
   # 运行时：--cap-drop=ALL --cap-add=NET_BIND_SERVICE
   

4. 自动化健康检查，赋予系统治愈的能力

   HEALTHCHECK --interval=30s --timeout=3s --start-period=10s --retries=3 \
     CMD curl -f http://localhost:8080/health || exit 1
   
   # 或使用自定义脚本
   HEALTHCHECK --interval=30s --timeout=5s --retries=3 \
     CMD /app/scripts/healthcheck.sh
   

5. 凡事都要明确，清晰的声明工作目录、暴露的端口和入口点

   # 明确的基础镜像版本
   FROM python:3.12-slim
   
   # 明确的工作目录
   WORKDIR /app
   
   # 明确的环境变量
   ENV PYTHONUNBUFFERED=1 \
       PYTHONDONTWRITEBYTECODE=1 \
       PATH="/app/bin:${PATH}"
   
   # 明确的暴露端口
   EXPOSE 8080/tcp
   
   # 明确的卷挂载点
   VOLUME ["/app/data", "/app/logs"]
   
   # 明确的入口点
   ENTRYPOINT ["python", "-m", "myapp"]
   CMD ["--config", "/app/config/config.yaml"]
   

Quick Reference

Dockerfile Best Practices

1. Use multi-stage builds for size reduction

# syntax=docker/dockerfile:1
FROM node:20 AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci
COPY . .
RUN npm run build

FROM node:20-slim
COPY --from=builder /app/dist ./dist
COPY --from=builder /app/node_modules ./node_modules
CMD ["node", "dist/server.js"]

Source: Official Docker documentation - Multi-stage builds

2. Choose minimal, trusted base images

# Good: Official slim images
FROM python:3.12-slim

# Better: Alpine-based (even smaller)
FROM python:3.12-alpine

# Best: distroless for production (minimal, auto-updates)
FROM gcr.io/distroless/python3-debian12

Source: Official Docker best practices

3. Combine RUN commands to reduce layers

# Bad: Multiple layers
RUN apt-get update
RUN apt-get install -y python3
RUN apt-get clean

# Good: Single layer with cleanup
RUN apt-get update && \
    apt-get install -y --no-install-recommends python3 && \
    rm -rf /var/lib/apt/lists/*

Source: Official Docker best practices

4. Use .dockerignore to exclude unnecessary files

# .dockerignore
node_modules
npm-debug.log
.git
.env.local
*.md
tests/

Source: Official Docker documentation

5. Run as non-root user

RUN adduser --disabled-password --gecos '' appuser
USER appuser

Source: Official Docker security documentation

6. Implement health checks

HEALTHCHECK --interval=30s --timeout=3s --start-period=5s --retries=3 \
  CMD curl -f http://localhost:8080/health || exit 1

Source: Official Docker documentation

7. Use specific versions, not latest

# Bad: Unpredictable updates
FROM ubuntu:latest

# Good: Reproducible builds
FROM ubuntu:24.04

Source: Official Docker best practices

Docker Compose Patterns

8. Environment variable interpolation

# compose.yaml
services:
  web:
    image: "webapp:${TAG:-latest}"
    environment:
      - DATABASE_URL=${DATABASE_URL}
      - DEBUG=${DEBUG:-false}

# .env
TAG=v1.5
DATABASE_URL=postgres://db:5432/app

Source: Official Docker Compose documentation

9. Multi-environment configuration

# compose.base.yaml (common configuration)
services:
  web:
    build: .
    ports:
      - "8000:8000"

# compose.production.yaml (production overrides)
services:
  web:
    restart: always
    environment:
      - NODE_ENV=production
    deploy:
      resources:
        limits:
          cpus: '0.50'
          memory: 512M

# Development
docker compose up

# Production
docker compose -f compose.base.yaml -f compose.production.yaml up -d

Source: Official Docker Compose production guide

10. Volume persistence with named volumes

services:
  db:
    image: postgres:16
    volumes:
      - pgdata:/var/lib/postgresql/data
    environment:
      - POSTGRES_PASSWORD=${DB_PASSWORD}

volumes:
  pgdata:
    # Named volume persists across container restarts

Source: Official Docker volumes documentation

11. Using secrets in Compose

services:
  web:
    image: myapp:latest
    secrets:
      - db_password
    environment:
      - DB_PASSWORD_FILE=/run/secrets/db_password

secrets:
  db_password:
    file: ./secrets/db_password.txt

Source: Official Docker security documentation

Data Persistence Patterns

12. Volume for persistent data

# Create and use a named volume
docker volume create mydata
docker run -d --name app \
  --mount source=mydata,target=/app/data \
  myapp:latest

Source: Official Docker volumes documentation

13. Bind mount for development

# Sync local files into container
docker run -d --name dev \
  --mount type=bind,source=$(pwd)/src,target=/app/src \
  myapp:latest

Source: Official Docker documentation

14. Read-only volume for shared data

services:
  web:
    image: nginx:alpine
    volumes:
      - static_content:/usr/share/nginx/html:ro

volumes:
  static_content:

Source: Official Docker volumes documentation

Container Security

15. Drop unnecessary capabilities

docker run -d --name secure_app \
  --cap-drop=ALL \
  --cap-add=NET_BIND_SERVICE \
  myapp:latest

Source: Official Docker security documentation

16. Run with resource limits

docker run -d --name app \
  --memory="512m" \
  --memory-reservation="256m" \
  --cpus="0.5" \
  --pids-limit 100 \
  myapp:latest

Source: Official Docker run documentation

17. Run as non-root

docker run -d --name app \
  --user 1000:1000 \
  myapp:latest

Source: Official Docker security documentation

Production Deployment

18. Health check in docker run

docker run -d --name app \
  --health-cmd="curl -f http://localhost/health || exit 1" \
  --health-interval=30s \
  --health-timeout=3s \
  --health-retries=3 \
  myapp:latest

Source: Official Docker run documentation

19. Update and redeploy a service

# Rebuild and recreate without affecting dependencies
docker compose up -d --build web

Source: Official Docker Compose production guide

20. Backup and restore volumes

# Backup a volume
docker run --rm --volumes-from app_container \
  -v $(pwd):/backup \
  alpine tar czf /backup/backup.tar.gz /data

# Restore to new container
docker run --rm --volumes-from new_app_container \
  -v $(pwd):/backup \
  alpine tar xzf /backup/backup.tar.gz

Source: Official Docker volumes documentation

Reference Files

This skill includes comprehensive documentation in references/:

Core Documentation

• getting_started.md (4 pages, confidence: medium)

  • Dockerfile best practices
  • Multi-stage builds
  • Base image selection
  • Environment variable management
  • Rebuilding and caching strategies
  • .dockerignore usage

• dockerfile.md (2 pages, confidence: medium)

  • Multi-stage build syntax
  • Build stage naming and targeting
  • External image usage
  • BuildKit vs legacy builder differences
  • Complete Dockerfile instruction reference

• compose.md (5 pages, confidence: medium)

  • Environment variable interpolation
  • .env file configuration
  • Variable precedence rules
  • Volume management in Compose
  • Production deployment patterns

• security.md (1 page, confidence: medium)

  • Container security configuration
  • Linux capabilities management
  • User permissions
  • Resource constraints
  • Health check parameters

• other.md (1 page, confidence: medium)

  • Docker run command reference
  • Container lifecycle management
  • Networking basics
  • Filesystem mounts

Usage Guidance

Use these reference files when you need:

• Detailed syntax: Check the appropriate reference file for complete syntax
• Examples: Each reference contains real-world examples from official docs
• Troubleshooting: Reference files contain common issues and solutions
• Best practices: Official recommendations for production deployments

Working with This Skill

For Beginners

1. Start with the concepts: Understand the core principles before diving into syntax
2. Use the quick reference: The examples above cover the most common patterns
3. Check reference files: When you need detailed information about specific commands

For Intermediate Users

1. Study multi-stage builds: Essential for production-grade images
2. Learn Compose patterns: Critical for multi-service applications
3. Understand volumes: Data persistence is key to production deployments
4. Implement security: Never run production containers without security hardening

For Advanced Users

1. Optimize build caching: Structure Dockerfiles to maximize cache hits
2. Resource management: Set appropriate CPU, memory, and I/O limits
3. Multi-environment configs: Use Compose overrides for dev/staging/prod
4. Health monitoring: Implement comprehensive health checks and logging
5. Secrets management: Use proper secret handling, never commit credentials

Common Workflows

Building a production image:

1. Start with a minimal base image
2. Use multi-stage builds to reduce size
3. Combine RUN commands to minimize layers
4. Add health checks
5. Run as non-root user
6. Test locally before pushing

Deploying with Compose:

1. Create base compose.yaml with service definitions
2. Use .env for environment-specific values
3. Create override files (e.g., compose.production.yaml) for different environments
4. Use named volumes for persistent data
5. Never commit secrets to version control

Managing data:

1. Use named volumes for persistent data
2. Use bind mounts for development (hot reload)
3. Implement backup strategies for volumes
4. Document volume contents and purpose

Notes

• Source: Official Docker documentation (docs.docker.com)
• Coverage: Dockerfiles, Docker Compose, volumes, security, and deployment
• Best practices: All examples follow official Docker recommendations
• Multi-environment: Patterns for dev, staging, and production
• Security: Emphasis on secure configurations and secrets management

Key Takeaways

1. Separate concerns: Dockerfiles (build), Compose (orchestration), .env (configuration)
2. Multi-stage builds: Use them to create small, secure production images
3. Never commit secrets: Use environment variables and secret management
4. Containers are ephemeral: Design with data persistence in mind
5. Security by default: Minimal images, non-root users, dropped capabilities
6. Reproducibility: Pin image versions, use specific tags, document dependencies
7. Health checks: Always implement health checks for production services
8. Resource limits: Set CPU, memory, and I/O limits to prevent runaway containers

Updating

To refresh this skill with updated documentation:

1. Re-run the scraper with the same configuration
2. The skill will be rebuilt with the latest information from official Docker documentation