wecom-ai-assistant/.github/workflows/deploy.yml

# 生产环境自动部署
# 触发条件：push 到 main 分支
# 功能：构建 backend 镜像 → 推送到 GHCR → SSH 部署到云服务器 → 健康检查

name: Deploy to Production

on:
  push:
    branches: [main]
  workflow_dispatch:
    inputs:
      image_tag:
        description: '镜像标签（默认: latest）'
        required: false
        default: 'latest'

env:
  REGISTRY: ghcr.io
  IMAGE_NAME_BACKEND: wecom-ai-backend

jobs:
  # 构建并推送 Backend 镜像
  build-backend:
    runs-on: ubuntu-latest
    outputs:
      image_tag: ${{ steps.meta.outputs.tags }}
      image_digest: ${{ steps.build.outputs.digest }}
    steps:
      - name: Checkout code
        uses: actions/checkout@v4

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3

      - name: Log in to Container Registry
        uses: docker/login-action@v3
        with:
          registry: ${{ env.REGISTRY }}
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}

      - name: Extract metadata
        id: meta
        uses: docker/metadata-action@v5
        with:
          images: ${{ env.REGISTRY }}/${{ github.repository_owner }}/${{ env.IMAGE_NAME_BACKEND }}
          tags: |
            type=raw,value=latest
            type=sha,prefix={{branch}}-
            type=sha,format=short

      - name: Build and push backend image
        id: build
        uses: docker/build-push-action@v6
        with:
          context: .
          file: ./deploy/docker/backend.Dockerfile
          push: true
          tags: ${{ steps.meta.outputs.tags }}
          labels: ${{ steps.meta.outputs.labels }}
          cache-from: type=gha
          cache-to: type=gha,mode=max

  # 部署到生产服务器
  deploy:
    runs-on: ubuntu-latest
    needs: build-backend
    if: github.ref == 'refs/heads/main'
    steps:
      - name: Checkout code
        uses: actions/checkout@v4

      - name: Set deployment variables
        id: vars
        run: |
          if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
            echo "image_tag=${{ github.event.inputs.image_tag }}" >> $GITHUB_OUTPUT
          else
            echo "image_tag=latest" >> $GITHUB_OUTPUT
          fi
          echo "domain=${{ secrets.PROD_DOMAIN }}" >> $GITHUB_OUTPUT

      - name: Prepare deployment token
        id: prepare-token
        run: |
          if [ -n "${{ secrets.GHCR_TOKEN }}" ]; then
            echo "token=${{ secrets.GHCR_TOKEN }}" >> $GITHUB_OUTPUT
          else
            echo "token=${{ secrets.GITHUB_TOKEN }}" >> $GITHUB_OUTPUT
          fi

      - name: Deploy to production server
        uses: appleboy/ssh-action@v1.0.0
        env:
          DEPLOY_TOKEN: ${{ steps.prepare-token.outputs.token }}
        with:
          host: ${{ secrets.PROD_HOST }}
          username: ${{ secrets.PROD_USER }}
          key: ${{ secrets.PROD_SSH_KEY }}
          port: ${{ secrets.PROD_SSH_PORT || 22 }}
          script: |
            set -e
            
            # 设置变量
            IMAGE_TAG="${{ steps.vars.outputs.image_tag }}"
            GITHUB_REPOSITORY_OWNER="${{ github.repository_owner }}"
            REGISTRY="${{ env.REGISTRY }}"
            IMAGE_NAME_BACKEND="${{ env.IMAGE_NAME_BACKEND }}"
            APP_PATH="${{ secrets.PROD_APP_PATH || '/opt/wecom-ai-assistant' }}"
            
            # 进入项目目录
            cd "${APP_PATH}" || { echo "错误: 无法进入目录 ${APP_PATH}"; exit 1; }
            
            # 登录到容器镜像仓库
            echo "${DEPLOY_TOKEN}" | docker login "${REGISTRY}" -u "${{ github.actor }}" --password-stdin || {
              echo "警告: Docker 登录失败，尝试继续部署（可能使用本地镜像）"
            }
            
            # 拉取最新镜像
            IMAGE_FULL="${REGISTRY}/${GITHUB_REPOSITORY_OWNER}/${IMAGE_NAME_BACKEND}:${IMAGE_TAG}"
            echo "拉取镜像: ${IMAGE_FULL}"
            docker pull "${IMAGE_FULL}" || {
              echo "警告: 拉取镜像失败，尝试使用 latest 标签"
              docker pull "${REGISTRY}/${GITHUB_REPOSITORY_OWNER}/${IMAGE_NAME_BACKEND}:latest"
              IMAGE_TAG="latest"
            }
            
            # 设置环境变量供 docker-compose 使用
            export IMAGE_TAG="${IMAGE_TAG}"
            export GITHUB_REPOSITORY_OWNER="${GITHUB_REPOSITORY_OWNER}"
            
            # 更新服务
            echo "更新服务..."
            docker compose -f docker-compose.prod.yml --env-file .env.prod pull backend 2>/dev/null || true
            docker compose -f docker-compose.prod.yml --env-file .env.prod up -d --force-recreate backend
            
            # 等待服务启动
            echo "等待服务启动..."
            sleep 15
            
            # 检查服务状态
            echo "服务状态:"
            docker compose -f docker-compose.prod.yml ps
            
            # 检查后端健康状态
            echo "检查后端健康状态..."
            for i in {1..10}; do
              if docker compose -f docker-compose.prod.yml exec -T backend python -c "import urllib.request; urllib.request.urlopen('http://localhost:8000/api/health')" 2>/dev/null; then
                echo "✓ 后端服务健康"
                break
              fi
              if [ $i -eq 10 ]; then
                echo "⚠ 后端服务可能未就绪，请检查日志"
              else
                sleep 2
              fi
            done

      - name: Health check
        run: |
          DOMAIN=${{ secrets.PROD_DOMAIN }}
          MAX_RETRIES=30
          RETRY_COUNT=0
          
          echo "健康检查: https://${DOMAIN}/api/health"
          
          while [ $RETRY_COUNT -lt $MAX_RETRIES ]; do
            if curl -f -s "https://${DOMAIN}/api/health" > /dev/null 2>&1; then
              echo "✓ 健康检查通过"
              exit 0
            fi
            RETRY_COUNT=$((RETRY_COUNT + 1))
            echo "等待服务就绪... ($RETRY_COUNT/$MAX_RETRIES)"
            sleep 2
          done
          
          echo "✗ 健康检查失败"
          exit 1