492 lines
11 KiB
Markdown
492 lines
11 KiB
Markdown
# Dockerfile 编写指南
|
||
|
||
> Dockerfile 是**镜像构建的蓝图**,每一行指令都会在镜像中生成一层。写好 Dockerfile 的核心目标:**镜像更小、构建更快、启动更安全**。
|
||
|
||
---
|
||
|
||
## 一、Dockerfile 基本结构
|
||
|
||
```dockerfile
|
||
# 1. 基础镜像
|
||
FROM node:20-slim AS builder
|
||
|
||
# 2. 维护者信息(可选)
|
||
LABEL maintainer="yourname@example.com"
|
||
|
||
# 3. 设置工作目录
|
||
WORKDIR /app
|
||
|
||
# 4. 复制依赖文件并安装
|
||
COPY package*.json ./
|
||
RUN npm ci --production
|
||
|
||
# 5. 复制源代码
|
||
COPY . .
|
||
|
||
# 6. 构建(开发环境需要,生产环境不需要)
|
||
RUN npm run build
|
||
|
||
# 7. 暴露端口
|
||
EXPOSE 3000
|
||
|
||
# 8. 启动命令
|
||
CMD ["node", "dist/index.js"]
|
||
```
|
||
|
||
---
|
||
|
||
## 二、常用指令详解
|
||
|
||
### 2.1 `FROM` — 基础镜像
|
||
|
||
```dockerfile
|
||
# 选镜像的通用原则:
|
||
# 有 Alpine 选 Alpine(体积小)
|
||
# 有 Slim 选 Slim(平衡体积和功能)
|
||
# 没有再选完整版本
|
||
|
||
FROM node:20-alpine # 推荐:最小体积
|
||
FROM node:20-slim # 推荐:平衡方案
|
||
FROM node:20 # 不推荐:包含完整 OS,镜像臃肿
|
||
FROM python:3.12-slim-bookworm
|
||
FROM nginx:alpine
|
||
FROM golang:1.22-alpine
|
||
FROM ubuntu:24.04 # 一般作为多阶段构建的第一步
|
||
```
|
||
|
||
**常见基础镜像体积对比:**
|
||
|
||
| 镜像 | 大小 | 适用场景 |
|
||
|------|------|----------|
|
||
| `alpine` | 5-10MB | 生产环境、追求极致体积 |
|
||
| `slim` | 100-200MB | 大多数生产场景,兼容性好 |
|
||
| 完整版 | 1GB+ | 开发环境、需要完整工具链 |
|
||
|
||
### 2.2 `RUN` — 执行命令
|
||
|
||
```dockerfile
|
||
# 合并 RUN 指令减少镜像层数
|
||
# 不推荐:每行一个 RUN,产生多层
|
||
RUN apt-get update
|
||
RUN apt-get install -y nginx
|
||
RUN rm -rf /var/lib/apt/lists/*
|
||
|
||
# 推荐:合并为一行,并用 && 连接,用 \ 换行
|
||
RUN apt-get update && \
|
||
apt-get install -y nginx && \
|
||
rm -rf /var/lib/apt/lists/*
|
||
|
||
# 注意:RUN 产生的层会**永久留在镜像中**,即使后续 rm 了也没用
|
||
# 所以更新和安装、清理要放在同一个 RUN 里
|
||
```
|
||
|
||
### 2.3 `COPY` vs `ADD`
|
||
|
||
```dockerfile
|
||
# COPY — 简单复制,推荐优先使用
|
||
COPY package.json ./
|
||
COPY . .
|
||
|
||
# ADD — 有额外功能但不可预测,不推荐
|
||
ADD https://example.com/file.tar.gz /tmp/ # 能下载远程文件,但容易出错
|
||
ADD file.tar.gz /app/ # 能自动解压,但 COPY + RUN tar 更清晰
|
||
|
||
# 结论:99% 的情况用 COPY,ADD 只在自动解压 tar 和下载远程文件时有用
|
||
```
|
||
|
||
### 2.4 `WORKDIR` — 工作目录
|
||
|
||
```dockerfile
|
||
WORKDIR /app # 设置工作目录
|
||
COPY package.json ./ # 相对于 /app
|
||
RUN npm install
|
||
COPY . . # 复制代码到 /app
|
||
|
||
# WORKDIR 会被后续指令继承,且会自动创建(不需要提前 mkdir)
|
||
WORKDIR /subdir # 现在是 /app/subdir
|
||
RUN pwd # 输出 /app/subdir
|
||
```
|
||
|
||
### 2.5 `EXPOSE` — 声明端口
|
||
|
||
```dockerfile
|
||
# 声明容器运行时监听的端口(文档性质,不自动映射)
|
||
EXPOSE 3000
|
||
EXPOSE 80 443
|
||
|
||
# 实际端口映射需要在 docker run -p 或 docker-compose 中指定
|
||
```
|
||
|
||
### 2.6 `ENV` — 环境变量
|
||
|
||
```dockerfile
|
||
ENV NODE_ENV=production
|
||
ENV APP_PORT=3000
|
||
ENV DB_HOST=localhost
|
||
|
||
# 容器启动时可以覆盖:
|
||
# docker run -e DB_HOST=192.168.1.100 my-app
|
||
```
|
||
|
||
### 2.7 `CMD` vs `ENTRYPOINT`
|
||
|
||
```dockerfile
|
||
# CMD — 容器启动时执行的默认命令(可被覆盖)
|
||
CMD ["node", "dist/index.js"]
|
||
# 运行 docker run my-app echo "hello" 会覆盖 CMD
|
||
|
||
# ENTRYPOINT — 入口点,不可被 docker run 覆盖(但可用 --entrypoint 强制覆盖)
|
||
ENTRYPOINT ["node"]
|
||
CMD ["dist/index.js"] # 作为 ENTRYPOINT 的参数
|
||
|
||
# 实际开发中,大多数情况用 CMD 就够了
|
||
```
|
||
|
||
---
|
||
|
||
## 三、多阶段构建(关键优化手段)
|
||
|
||
### 3.1 为什么需要多阶段构建
|
||
|
||
前端项目用 Next.js/Vite + TypeScript 构建,需要 Node.js 全套环境。但运行只需要一个编译后的静态文件。多阶段构建可以**把构建工具留在第一阶段,最终镜像只放产物**。
|
||
|
||
### 3.2 前端多阶段构建(Node + TypeScript)
|
||
|
||
```dockerfile
|
||
# ========== 第一阶段:构建 ==========
|
||
FROM node:20-alpine AS builder
|
||
|
||
WORKDIR /app
|
||
|
||
# ① 先复制依赖文件,利用 Docker 缓存
|
||
COPY package*.json ./
|
||
RUN npm ci
|
||
|
||
# ② 再复制源码(代码改了才重新构建)
|
||
COPY . .
|
||
|
||
# ③ 构建
|
||
RUN npm run build
|
||
|
||
# ========== 第二阶段:运行 ==========
|
||
FROM node:20-alpine AS production
|
||
|
||
# 安装 tini(处理僵尸进程,可选但推荐)
|
||
RUN apk add --no-cache tini
|
||
|
||
WORKDIR /app
|
||
|
||
# 仅复制 production 依赖
|
||
COPY package*.json ./
|
||
RUN npm ci --production
|
||
|
||
# 从构建阶段复制产物
|
||
COPY --from=builder /app/dist ./dist
|
||
COPY --from=builder /app/public ./public
|
||
|
||
ENV NODE_ENV=production
|
||
|
||
EXPOSE 3000
|
||
|
||
# 使用 tini 作为 PID 1
|
||
ENTRYPOINT ["/sbin/tini", "--"]
|
||
CMD ["node", "dist/index.js"]
|
||
```
|
||
|
||
**核心思路:先用大镜像构建 → 再把产物复制到最小镜像中运行。**
|
||
|
||
### 3.3 Java 多阶段构建
|
||
|
||
```dockerfile
|
||
# 第一阶段:Maven 构建
|
||
FROM maven:3.9-eclipse-temurin-21 AS build
|
||
WORKDIR /app
|
||
COPY pom.xml .
|
||
RUN mvn dependency:go-offline -B
|
||
COPY src ./src
|
||
RUN mvn package -DskipTests
|
||
|
||
# 第二阶段:运行
|
||
FROM eclipse-temurin:21-jre-alpine
|
||
WORKDIR /app
|
||
COPY --from=build /app/target/*.jar app.jar
|
||
EXPOSE 8080
|
||
ENTRYPOINT ["java", "-jar", "app.jar"]
|
||
```
|
||
|
||
### 3.4 Go 多阶段构建
|
||
|
||
```dockerfile
|
||
# 第一阶段:编译
|
||
FROM golang:1.22-alpine AS builder
|
||
WORKDIR /app
|
||
COPY go.mod go.sum ./
|
||
RUN go mod download
|
||
COPY . .
|
||
RUN CGO_ENABLED=0 go build -o /app/server
|
||
|
||
# 第二阶段:运行
|
||
FROM alpine:3.19
|
||
RUN apk --no-cache add ca-certificates
|
||
COPY --from=builder /app/server /app/server
|
||
EXPOSE 8080
|
||
CMD ["/app/server"]
|
||
```
|
||
|
||
---
|
||
|
||
## 四、.dockerignore 文件
|
||
|
||
> **和 `.gitignore` 一样重要!** 不写 `.dockerignore` 会导致不必要的文件被复制到镜像中,增大体积、暴露敏感信息。
|
||
|
||
```dockerignore
|
||
# 依赖
|
||
node_modules
|
||
npm-debug.log
|
||
|
||
# Git
|
||
.git
|
||
.gitignore
|
||
|
||
# 构建产物(别把 build 直接拷进去,让容器自己 build)
|
||
dist
|
||
build
|
||
*.local
|
||
|
||
# 编辑器
|
||
.vscode
|
||
.idea
|
||
*.swp
|
||
|
||
# Docker 自身
|
||
docker-compose.yml
|
||
Dockerfile
|
||
.dockerignore
|
||
|
||
# 文档和测试
|
||
README.md
|
||
LICENSE
|
||
tests
|
||
__tests__
|
||
|
||
# 环境变量(千万不要拷进去!)
|
||
.env
|
||
.env.local
|
||
.env.*.local
|
||
```
|
||
|
||
---
|
||
|
||
## 五、最佳实践清单
|
||
|
||
### ✅ 应该做
|
||
|
||
| 实践 | 说明 |
|
||
|------|------|
|
||
| 使用具体版本号 | `FROM node:20.11-alpine`,不要用 `latest` |
|
||
| 用 `alpine` 或 `slim` | 减小镜像体积 |
|
||
| 合并 `RUN` 指令 | 减少镜像层数 |
|
||
| 先复制 `package.json` 再 `npm install` | 利用 Docker 层缓存 |
|
||
| 用 `.dockerignore` | 排除不必要的文件 |
|
||
| 用多阶段构建 | 分离构建和运行环境 |
|
||
| 非 root 用户运行 | 安全考虑 |
|
||
| 用 `CMD ["exec格式"]` | 避免 shell 包装问题 |
|
||
|
||
### ❌ 不应该做
|
||
|
||
| 实践 | 问题 |
|
||
|------|------|
|
||
| 用 `latest` 标签 | 构建结果不可复现 |
|
||
| 在 Dockerfile 中存密码/密钥 | 暴露敏感信息 |
|
||
| 用 `ADD` 替代 `COPY` | 不可预测行为 |
|
||
| 安装不必要的应用 | 镜像臃肿 |
|
||
| 用 `RUN apt-get install && apt-get remove` | 删除的文件仍在镜像层中 |
|
||
| 以 root 用户运行应用 | 安全风险 |
|
||
|
||
---
|
||
|
||
## 六、非 Root 用户示例
|
||
|
||
```dockerfile
|
||
FROM node:20-alpine
|
||
|
||
WORKDIR /app
|
||
|
||
COPY package*.json ./
|
||
RUN npm ci
|
||
|
||
COPY . .
|
||
|
||
# 创建非 root 用户
|
||
RUN addgroup -S appgroup && \
|
||
adduser -S appuser -G appgroup
|
||
|
||
# 切换用户
|
||
USER appuser
|
||
|
||
EXPOSE 3000
|
||
|
||
CMD ["node", "dist/index.js"]
|
||
```
|
||
|
||
---
|
||
|
||
## 七、常见问题排查
|
||
|
||
### 7.1 镜像太大
|
||
|
||
```bash
|
||
# 分析镜像各层大小
|
||
docker history my-app:latest
|
||
|
||
# 检查是否有未清理的缓存
|
||
# 解决:使用 alpine/slim 基础镜像 + 多阶段构建 + .dockerignore
|
||
```
|
||
|
||
### 7.2 构建缓存失效
|
||
|
||
```dockerfile
|
||
# 错误:代码改动导致 npm install 也重新执行
|
||
COPY . .
|
||
RUN npm install
|
||
|
||
# 正确:依赖文件不变时跳过 npm install
|
||
COPY package*.json ./
|
||
RUN npm install
|
||
COPY . .
|
||
```
|
||
|
||
### 7.3 容器启动后立即退出
|
||
|
||
```bash
|
||
# 查看日志
|
||
docker logs <容器名>
|
||
|
||
# 常见原因:
|
||
# 1. CMD/ENTRYPOINT 写错了
|
||
# 2. 主进程不是 PID 1(用 tini 或 dumb-init 处理)
|
||
# 3. 应用启动失败(检查日志)
|
||
```
|
||
|
||
### 7.4 端口无法访问
|
||
|
||
```bash
|
||
# 1. 检查 docker run -p 是否映射
|
||
docker run -p 3000:3000 ...
|
||
|
||
# 2. 检查容器内应用是否监听 0.0.0.0(不是 localhost)
|
||
# 3. 检查防火墙
|
||
|
||
# 4. 用 docker-compose 时,确认 services 定义正确
|
||
```
|
||
|
||
---
|
||
|
||
## 七、Docker 安全加固
|
||
|
||
### 7.1 使用非 root 用户
|
||
|
||
```dockerfile
|
||
# 最简方式(Node 官方镜像已内置 node 用户)
|
||
FROM node:20-alpine
|
||
USER node
|
||
CMD ["node", "server.js"]
|
||
|
||
# 手动创建用户(通用模式)
|
||
FROM node:20-alpine
|
||
WORKDIR /app
|
||
COPY --chown=node:node package*.json ./
|
||
RUN npm ci
|
||
COPY --chown=node:node . .
|
||
USER node
|
||
CMD ["node", "dist/index.js"]
|
||
```
|
||
|
||
> `--chown=user:group` 在 COPY/RUN 时设置文件所有者,**避免 root 拥有文件**后切换到非 root 用户导致权限问题。
|
||
|
||
### 7.2 不要安装 build 依赖到生产镜像
|
||
|
||
```dockerfile
|
||
# 错误:所有依赖都装进去了
|
||
FROM node:20-alpine
|
||
RUN npm install # devDependencies 也被安装了
|
||
|
||
# 正确:生产环境只装 dependencies
|
||
RUN npm ci --only=production
|
||
|
||
# 更正确:多阶段构建
|
||
FROM node:20-alpine AS builder
|
||
RUN npm install # 包含 devDependencies
|
||
RUN npm run build
|
||
|
||
FROM node:20-alpine
|
||
RUN npm ci --only=production
|
||
COPY --from=builder /app/dist ./dist
|
||
```
|
||
|
||
### 7.3 扫描镜像漏洞
|
||
|
||
```bash
|
||
# 使用 trivy(推荐)
|
||
docker run --rm -v /var/run/docker.sock:/var/run/docker.sock \
|
||
aquasec/trivy image nginx:latest
|
||
|
||
# 扫描指定镜像
|
||
trivy image my-app:v1.0
|
||
|
||
# 只输出高危及以上
|
||
trivy image --severity HIGH,CRITICAL my-app:v1.0
|
||
|
||
# 扫描 Dockerfile 本身(不用构建镜像)
|
||
trivy config Dockerfile
|
||
```
|
||
|
||
### 7.4 密钥管理原则
|
||
|
||
```dockerfile
|
||
# ❌ 绝对不要硬编码密码
|
||
RUN echo "password=123456" > /app/config.env
|
||
|
||
# ❌ 不要在构建时传递密钥(会留在镜像层中)
|
||
docker build --build-arg DB_PASSWORD=secret .
|
||
|
||
# ✅ 运行时传入环境变量
|
||
docker run -e DB_PASSWORD=secret my-app
|
||
|
||
# ✅ 或使用 Docker secrets(Swarm 模式)
|
||
# ✅ 生产环境用 Kubernetes Secrets / AWS Secrets Manager
|
||
```
|
||
|
||
---
|
||
|
||
## 八、实战:Node.js + Express 完整 Dockerfile
|
||
|
||
```dockerfile
|
||
# Build stage
|
||
FROM node:20-alpine AS builder
|
||
WORKDIR /app
|
||
COPY package*.json ./
|
||
RUN npm ci --only=production
|
||
COPY . .
|
||
|
||
# Production stage
|
||
FROM node:20-alpine
|
||
WORKDIR /app
|
||
COPY --from=builder /app/node_modules ./node_modules
|
||
COPY --from=builder /app/dist ./dist
|
||
COPY --from=builder /app/package.json ./package.json
|
||
|
||
ENV NODE_ENV=production
|
||
ENV PORT=3000
|
||
|
||
RUN addgroup -S app && adduser -S app -G app
|
||
USER app
|
||
|
||
EXPOSE 3000
|
||
CMD ["node", "dist/server.js"]
|
||
```
|
||
|
||
---
|
||
|
||
> 下一步:理解 Dockerfile 后,学习 [[docker-compose]] 来编排多容器应用。
|