GitHub Actions中用环境变量指定Runner组失败，求可行方案

问题根源 #

GitHub Actions中runs-on字段的解析时机早于environment环境变量的加载，所以直接在runs-on里引用环境级vars会返回空值；而复用工作流时给调用job指定environment会触发GitHub的限制，导致job无法正常传递参数。

可行解决方案 #

方案1：矩阵策略+环境变量映射

预先在仓库变量中按环境定义对应的Runner组、标签等配置，通过矩阵动态匹配用户选择的环境：

name: 切换维护模式

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: false

on:
  workflow_dispatch:
    inputs:
      environment:
        type: choice
        options:
          - Acceptance
          - Production
        required: true
      maintenance-text:
        description: '维护模式显示文本'
        required: true

jobs:
  maintenance-mode-toggle:
    name: 执行维护模式切换
    strategy:
      matrix:
        include:
          - env: Acceptance
            runner-group: ${{ vars.ACCEPTANCE_RUNNER_GROUP }}
            iis-label: ${{ vars.ACCEPTANCE_IIS_LABEL }}
            iis-dir: ${{ vars.ACCEPTANCE_IIS_DIR }}
          - env: Production
            runner-group: ${{ vars.PRODUCTION_RUNNER_GROUP }}
            iis-label: ${{ vars.PRODUCTION_IIS_LABEL }}
            iis-dir: ${{ vars.PRODUCTION_IIS_DIR }}
    if: ${{ matrix.env == inputs.environment }}
    environment: ${{ matrix.env }}
    runs-on:
      group: ${{ matrix.runner-group }}
      labels: ${{ matrix.iis-label }}
    steps:
      - name: IIS: 启用维护模式
        run: |  
          echo "${{ inputs.maintenance-text }}" > app_offline.htm;
        working-directory: ${{ matrix.iis-dir }}

操作说明：在仓库设置的「变量」中创建对应环境的变量（如ACCEPTANCE_RUNNER_GROUP），矩阵会在工作流初始化阶段解析这些仓库变量，避开environment加载时机的问题。

方案2：前置job获取环境变量后传递

用一个轻量前置job先加载目标环境、获取变量，再通过输出传递给实际执行的复用工作流：

name: Elog - 维护模式管理

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: false

on:
  workflow_dispatch:
    inputs:
      environment:
        type: choice
        options:
          - Acceptance
          - Production
        required: true
      toggle:
        type: choice
        options:
          - 'on'
          - 'off'
        required: true
      message:
        description: '维护页面显示消息'

jobs:
  get-runner-config:
    name: 获取Runner配置参数
    runs-on: ubuntu-latest
    environment: ${{ inputs.environment }}
    outputs:
      runner-group: ${{ steps.extract-vars.outputs.runner-group }}
      iis-label: ${{ steps.extract-vars.outputs.iis-label }}
      iis-dir: ${{ steps.extract-vars.outputs.iis-dir }}
    steps:
      - id: extract-vars
        run: |
          echo "runner-group=${{ vars.RUNNER_GROUP }}" >> $GITHUB_OUTPUT
          echo "iis-label=${{ vars.IIS_APPLICATION_LABEL }}" >> $GITHUB_OUTPUT
          echo "iis-dir=${{ vars.IIS_APPLICATION_DIRECTORY }}" >> $GITHUB_OUTPUT

  deploy-maintenance-mode:
    name: 执行维护模式切换
    if: ${{ github.ref == 'refs/heads/main' }}
    needs: get-runner-config
    uses: ./.github/workflows/maitenance-template-workflow.yml
    with:
      application-pool: default
      environment: ${{ inputs.environment }}
      toggle: ${{ inputs.toggle }}
      runner-group: ${{ needs.get-runner-config.outputs.runner-group }}
      iis-application-label: ${{ needs.get-runner-config.outputs.iis-label }}
      message: ${{ inputs.message }}

操作说明：前置job先加载指定环境拿到变量，再通过outputs传递给复用工作流，绕开了直接给复用job指定environment的限制。

方案3：按命名约定直接插值引用

如果环境变量是按「前缀+环境名」的规则命名（如RUNNER_GROUP_ACCEPTANCE、RUNNER_GROUP_PRODUCTION），可以直接通过变量插值匹配：

jobs:
  maintenance-mode-toggle:
    environment: ${{ inputs.environment }}
    runs-on:
      group: ${{ vars[format('RUNNER_GROUP_{0}', inputs.environment)] }}
      labels: ${{ vars[format('IIS_APPLICATION_LABEL_{0}', inputs.environment)] }}
    steps:
      - name: IIS: 启用维护模式
        run: |  
          echo "${{ inputs.maintenance-text }}" > app_offline.htm;
        working-directory: ${{ vars[format('IIS_APPLICATION_DIRECTORY_{0}', inputs.environment)] }}

操作说明：利用format函数拼接变量名，直接从仓库变量中匹配对应环境的配置，适合变量命名规范统一的场景。

内容的提问来源于stack exchange，提问作者Nick