如何在Github Actions跨仓库触发部署并等待状态完成后续操作？

核心逻辑 #

1. 依托后端仓库的Github Actions触发前端部署（你已通过Github API实现）
2. 轮询前端仓库的部署状态，直到部署进入完成态（成功/失败）
3. 前端部署成功后，执行全栈应用发布、初始测试及其他自动化操作

1. 等待前端部署状态的实现 #

在后端仓库的Github Actions Workflow中，新增轮询部署状态的步骤。可以直接调用Github的Deployment Statuses API，或者用官方gh CLI简化操作。

示例Workflow片段（使用gh CLI）

jobs:
  trigger-frontend-deployment:
    runs-on: ubuntu-latest
    steps:
      - name: 检出后端代码
        uses: actions/checkout@v4

      - name: 调用Github API触发前端部署
        run: |
          # 你已实现的触发逻辑，此处示例为创建前端部署记录
          DEPLOYMENT_ID=$(curl -X POST \
            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
            -H "Accept: application/vnd.github.v3+json" \
            https://api.github.com/repos/<你的用户名>/<前端仓库名>/deployments \
            -d '{"ref":"main","environment":"production","auto_merge":false}' | jq -r '.id')
          echo "DEPLOYMENT_ID=$DEPLOYMENT_ID" >> $GITHUB_ENV

      - name: 安装Github CLI
        run: |
          type -p curl >/dev/null || (sudo apt update && sudo apt install curl -y)
          curl -fsSL https://cli.github.com/packages/githubcli-archive-keyring.gpg | sudo dd of=/usr/share/keyrings/githubcli-archive-keyring.gpg
          sudo chmod go+r /usr/share/keyrings/githubcli-archive-keyring.gpg
          echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" | sudo tee /etc/apt/sources.list.d/github-cli.list > /dev/null
          sudo apt update && sudo apt install gh -y

      - name: 等待前端部署完成
        run: |
          STATUS="pending"
          while [ "$STATUS" = "pending" ]; do
            echo "等待前端部署完成..."
            sleep 30
            STATUS=$(gh api repos/<你的用户名>/<前端仓库名>/deployments/${{ env.DEPLOYMENT_ID }}/statuses --jq '.[0].state')
            if [ "$STATUS" = "failure" ]; then
              echo "前端部署失败！终止流程"
              exit 1
            fi
          done
          echo "前端部署成功！"
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

  # 部署完成后的全栈操作
  full-stack-post-deployment:
    runs-on: ubuntu-latest
    needs: trigger-frontend-deployment
    steps:
      - name: 检出后端代码
        uses: actions/checkout@v4

      - name: 部署后端到生产环境
        run: |
          # 替换为你的后端部署脚本，比如Capistrano、Docker部署命令
          bundle exec cap production deploy

      - name: 运行全栈集成测试
        run: |
          # 根据实际需求启动服务并执行测试
          rails server -d
          npm run test:e2e # 示例为前端E2E测试（如Cypress/Playwright）

      - name: 执行部署后自动化操作
        run: |
          # 替换为你的自动化逻辑，比如清理缓存、发送通知
          rails runner "CacheCleanupJob.perform_now"
          curl -X POST ${{ secrets.SLACK_WEBHOOK }} -d '{"text":"全栈应用部署完成"}'

关键注意事项 #

• 权限配置：确保GITHUB_TOKEN或个人访问令牌（PAT）拥有访问前端仓库部署状态的权限
• 轮询频率：调整sleep时长避免触发Github API速率限制
• 失败拦截：前端部署失败时立即终止后续流程，避免无效操作
• 流程依赖：通过needs关键字确保全栈操作仅在前端部署成功后执行

内容的提问来源于stack exchange，提问作者Obaid Ashfaq