Docker + Kubernetes 生产环境部署完全指南

在企业级应用开发中，容器化与编排技术已经成为基础设施的基石。本文将从实际生产场景出发，系统讲解 Docker 容器化到 Kubernetes 编排部署的完整工作流，帮助 DevOps 工程师和后端开发者构建可靠、高效的云原生应用交付体系。

一、Docker 容器化最佳实践

1.1 多阶段构建

多阶段构建是生产环境镜像优化的核心手段，能够将构建依赖与运行时环境严格分离，显著减小最终镜像体积。

# ===== 阶段一：构建阶段 =====
FROM node:20-alpine AS builder

WORKDIR /app

# 先复制依赖文件，利用 Docker 缓存层加速构建
COPY package.json package-lock.json ./
RUN npm ci --only=production

# 复制源码并构建
COPY . .
RUN npm run build

# ===== 阶段二：运行阶段 =====
FROM node:20-alpine AS runtime

# 创建非 root 用户
RUN addgroup -S appgroup && adduser -S appuser -G appgroup

WORKDIR /app

# 仅从构建阶段复制必要产物
COPY --from=builder /app/dist ./dist
COPY --from=builder /app/node_modules ./node_modules
COPY --from=builder /app/package.json ./

USER appuser

EXPOSE 3000

HEALTHCHECK --interval=30s --timeout=3s --start-period=5s --retries=3 \
  CMD wget --no-verbose --tries=1 --spider http://localhost:3000/health || exit 1

CMD ["node", "dist/server.js"]

1.2 镜像优化要点

选择精简基础镜像：优先使用 alpine、slim 等变体，避免安装不必要的系统包
合并 RUN 指令：减少镜像层数，将相关操作用 && 连接
利用构建缓存：将不常变化的文件（如依赖声明）放在 Dockerfile 前部
设置 .dockerignore：排除 .git、node_modules、*.md 等无关文件

# .dockerignore
.git
.gitignore
node_modules
npm-debug.log
README.md
.env.local

二、Docker Compose 开发环境配置

在开发阶段，Docker Compose 能够快速搭建包含应用、数据库、缓存等完整依赖的本地环境。

# docker-compose.yml
version: "3.9"

services:
  app:
    build:
      context: .
      dockerfile: Dockerfile.dev
    ports:
      - "3000:3000"
    environment:
      - NODE_ENV=development
      - DB_HOST=postgres
      - DB_PORT=5432
      - REDIS_HOST=redis
    volumes:
      - .:/app          # 挂载源码实现热重载
      - /app/node_modules
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy

  postgres:
    image: postgres:16-alpine
    environment:
      POSTGRES_DB: myapp
      POSTGRES_USER: appuser
      POSTGRES_PASSWORD: secretpassword
    ports:
      - "5432:5432"
    volumes:
      - pgdata:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U appuser -d myapp"]
      interval: 5s
      timeout: 3s
      retries: 5

  redis:
    image: redis:7-alpine
    ports:
      - "6379:6379"
    volumes:
      - redisdata:/data
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 5s
      timeout: 3s
      retries: 5

volumes:
  pgdata:
  redisdata:

三、Kubernetes 核心概念

在深入部署配置之前，先梳理 K8s 中最核心的几个资源对象：

资源对象	职责
Pod	最小调度单元，包含一个或多个共享网络与存储的容器
Deployment	声明式管理 Pod 副本数，支持滚动更新与回滚
Service	为 Pod 提供稳定的网络访问入口与负载均衡
ConfigMap	存储非敏感配置数据，以环境变量或卷挂载方式注入 Pod
Secret	存储敏感数据（密码、证书等），Base64 编码存储

四、生产级 Kubernetes 部署配置

以下是一个完整的生产级部署方案，包含 Deployment、Service、ConfigMap、Secret 和 Ingress。

4.1 Namespace 与 ConfigMap

# namespace.yaml
apiVersion: v1
kind: Namespace
metadata:
  name: production
  labels:
    env: production
    team: backend
---
# configmap.yaml
apiVersion: v1
kind: ConfigMap
metadata:
  name: app-config
  namespace: production
data:
  NODE_ENV: "production"
  DB_HOST: "postgres-service"
  DB_PORT: "5432"
  DB_NAME: "myapp"
  REDIS_HOST: "redis-service"
  LOG_LEVEL: "info"

4.2 Secret

# secret.yaml
apiVersion: v1
kind: Secret
metadata:
  name: app-secrets
  namespace: production
type: Opaque
data:
  # 值需经过 base64 编码
  DB_PASSWORD: c2VjcmV0cGFzc3dvcmQ=
  JWT_SECRET: eW91ci1qd3Qtc2VjcmV0LWtleS0yMDI2
  API_KEY: YWJjZGVmZzEyMzQ1Njc4

注意：生产环境中建议使用 Sealed Secrets、HashiCorp Vault 或云厂商的密钥管理服务来管理敏感数据，避免将 Secret 明文提交到代码仓库。

4.3 Deployment

# deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp
  namespace: production
  labels:
    app: myapp
    version: v1.2.0
spec:
  replicas: 3
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1          # 滚动更新时最多多出 1 个 Pod
      maxUnavailable: 0    # 更新期间不允许有 Pod 不可用
  selector:
    matchLabels:
      app: myapp
  template:
    metadata:
      labels:
        app: myapp
        version: v1.2.0
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "3000"
        prometheus.io/path: "/metrics"
    spec:
      serviceAccountName: myapp-sa
      terminationGracePeriodSeconds: 60

      # 反亲和性：确保 Pod 分散在不同节点
      affinity:
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 100
              podAffinityTerm:
                labelSelector:
                  matchExpressions:
                    - key: app
                      operator: In
                      values:
                        - myapp
                topologyKey: kubernetes.io/hostname

      containers:
        - name: myapp
          image: registry.example.com/myapp:v1.2.0
          ports:
            - containerPort: 3000
              protocol: TCP

          # 资源限制与请求
          resources:
            requests:
              cpu: "250m"
              memory: "256Mi"
            limits:
              cpu: "500m"
              memory: "512Mi"

          # 环境变量注入
          envFrom:
            - configMapRef:
                name: app-config
            - secretRef:
                name: app-secrets

          # 健康检查
          livenessProbe:
            httpGet:
              path: /health
              port: 3000
            initialDelaySeconds: 15
            periodSeconds: 20
            timeoutSeconds: 5
            failureThreshold: 3

          readinessProbe:
            httpGet:
              path: /ready
              port: 3000
            initialDelaySeconds: 5
            periodSeconds: 10
            timeoutSeconds: 3
            failureThreshold: 3

          startupProbe:
            httpGet:
              path: /health
              port: 3000
            initialDelaySeconds: 0
            periodSeconds: 5
            failureThreshold: 12   # 最多等待 60 秒启动

          # 生命周期钩子
          lifecycle:
            preStop:
              exec:
                command: ["/bin/sh", "-c", "sleep 15"]

          # 安全上下文
          securityContext:
            runAsNonRoot: true
            runAsUser: 1000
            readOnlyRootFilesystem: true
            allowPrivilegeEscalation: false

4.4 Service 与 Ingress

# service.yaml
apiVersion: v1
kind: Service
metadata:
  name: myapp-service
  namespace: production
spec:
  type: ClusterIP
  selector:
    app: myapp
  ports:
    - name: http
      port: 80
      targetPort: 3000
      protocol: TCP
---
# ingress.yaml
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: myapp-ingress
  namespace: production
  annotations:
    nginx.ingress.kubernetes.io/rewrite-target: /
    nginx.ingress.kubernetes.io/ssl-redirect: "true"
    nginx.ingress.kubernetes.io/rate-limit: "100"
    cert-manager.io/cluster-issuer: letsencrypt-prod
spec:
  ingressClassName: nginx
  tls:
    - hosts:
        - app.example.com
      secretName: app-tls-secret
  rules:
    - host: app.example.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: myapp-service
                port:
                  number: 80

五、健康检查与自动扩缩容

5.1 健康检查设计

K8s 提供三种探针（Probe），分别应对不同场景：

livenessProbe：检测容器是否存活，失败则重启容器
readinessProbe：检测容器是否就绪，失败则从 Service Endpoints 中摘除
startupProbe：检测容器是否启动完成，适用于启动较慢的应用

5.2 HPA 自动扩缩容

# hpa.yaml
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: myapp-hpa
  namespace: production
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: myapp
  minReplicas: 3
  maxReplicas: 15
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70
    - type: Resource
      resource:
        name: memory
        target:
          type: Utilization
          averageUtilization: 80
  behavior:
    scaleUp:
      stabilizationWindowSeconds: 60
      policies:
        - type: Percent
          value: 50
          periodSeconds: 60
    scaleDown:
      stabilizationWindowSeconds: 300
      policies:
        - type: Percent
          value: 10
          periodSeconds: 60

HPA 的 behavior 配置至关重要：扩容可以激进一些以快速响应流量增长，缩容则应保守以避免频繁波动。

六、滚动更新与回滚策略

6.1 滚动更新

K8s Deployment 默认采用滚动更新策略。在上述 Deployment 配置中，我们设置了 maxSurge: 1 和 maxUnavailable: 0，这意味着更新时会先创建新 Pod，确认其就绪后再终止旧 Pod，实现零停机部署。

# 执行滚动更新
kubectl set image deployment/myapp \
  myapp=registry.example.com/myapp:v1.3.0 \
  -n production

# 查看更新状态
kubectl rollout status deployment/myapp -n production

# 查看更新历史
kubectl rollout history deployment/myapp -n production

6.2 回滚

当新版本出现问题时，可以快速回滚到上一个稳定版本：

# 回滚到上一版本
kubectl rollout undo deployment/myapp -n production

# 回滚到指定版本
kubectl rollout undo deployment/myapp --to-revision=2 -n production

# 查看当前版本
kubectl rollout history deployment/myapp -n production

七、监控与日志

7.1 Prometheus + Grafana 监控栈

在 Deployment 中已通过注解暴露了 /metrics 端点。以下是 Prometheus 的 ServiceMonitor 配置：

# servicemonitor.yaml
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: myapp-monitor
  namespace: production
  labels:
    release: prometheus
spec:
  selector:
    matchLabels:
      app: myapp
  endpoints:
    - port: http
      path: /metrics
      interval: 15s
      scrapeTimeout: 10s

7.2 日志收集

生产环境推荐使用 EFK（Elasticsearch + Fluentd + Kibana）或 PLG（Promtail + Loki + Grafana）日志栈。以下是一个 Fluentd DaemonSet 的核心配置片段：


yaml
# fluentd-configmap.yaml（节选）
apiVersion: v1
kind: ConfigMap
metadata:

---

> 📢 **本文为精简版，完整版包含独家工具推荐和深度分析，请访问 [WD Tech Blog](https://wdsega.github.io) 查看！**

*关注我的博客获取最新科技资讯、AI教程和效率工具推荐！*

推荐订阅源

DEV Community