e2e-test.yml

name: E2E Test
on:
  workflow_dispatch:

jobs:
  # Launch Lambda instance and register runner
  lambda:
    uses: ./.github/workflows/runner.yml
    secrets: inherit
    with:
      instance_type: gpu_1x_a10
      region: us-east-1
      runner_grace_period: "30"
      runner_initial_grace_period: "120"
      debug: "true"

  # Run test job on the Lambda runner
  gpu-test:
    needs: lambda
    runs-on: ${{ needs.lambda.outputs.id }}
    steps:
      - name: System info
        run: |
          echo "=== System Info ==="
          echo "Hostname: $(hostname)"
          echo "OS: $(cat /etc/os-release | grep PRETTY_NAME)"
          echo "Kernel: $(uname -r)"
          echo "User: $(whoami)"

      - name: GPU info
        run: |
          echo "=== GPU Info ==="
          nvidia-smi

      - name: CUDA check
        run: |
          echo "=== CUDA Version ==="
          nvcc --version || echo "nvcc not in PATH"
          cat /usr/local/cuda/version.txt 2>/dev/null || echo "CUDA version file not found"

      - name: Success
        run: echo "E2E test completed successfully on Lambda Labs GPU instance!"