docker.yml

---
name: Docker Build + Full Scale Test

on:
  push:
    tags:
      - 'v*'
  workflow_dispatch:

env:
  REGISTRY: ghcr.io
  IMAGE_NAME: ${{ github.repository }}

jobs:
  docker-build:
    name: Build + Full Scale Test (1000 Sessions)
    runs-on: ubuntu-24.04
    timeout-minutes: 45
    permissions:
      contents: read
      packages: write
    outputs:
      image: ${{ steps.meta.outputs.tags }}
      digest: ${{ steps.build.outputs.digest }}
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
      - name: Set up Rust toolchain
        uses: dtolnay/rust-toolchain@stable
      - name: Set up sccache
        uses: mozilla-actions/sccache-action@v0.0.9
      - name: Run internal scale test (1000 sessions)
        run: |
          cargo test --package phantom-mcp --test scale_test scale_full_1000 \
            -- --nocapture --test-threads=1
      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3
      - name: Log in to GHCR
        uses: docker/login-action@v3
        with:
          registry: ${{ env.REGISTRY }}
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Extract image metadata
        id: meta
        uses: docker/metadata-action@v5
        with:
          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
          tags: |
            type=ref,event=branch
            type=sha,prefix=sha-
            type=raw,value=latest,enable={{is_default_branch}}
      - name: Build image
        id: build
        uses: docker/build-push-action@v6
        with:
          context: .
          file: Dockerfile
          push: ${{ startsWith(github.ref, 'refs/tags/v') }}
          load: true
          tags: ${{ steps.meta.outputs.tags }}
          labels: ${{ steps.meta.outputs.labels }}
          cache_from: type=gha
          cache_to: type=gha,mode=max
          build-args: |
            BUILDKIT_INLINE_CACHE=1
          platforms: linux/amd64
      - name: Resolve API key for compose + k6
        env:
          PHANTOM_API_KEY_SECRET: ${{ secrets.PHANTOM_API_KEY }}
        run: |
          if [ -n "${PHANTOM_API_KEY_SECRET}" ]; then
            echo "PHANTOM_API_KEY=${PHANTOM_API_KEY_SECRET}" >> "$GITHUB_ENV"
          else
            echo "PHANTOM_API_KEY=phantom-ci-key" >> "$GITHUB_ENV"
          fi
          echo "PHANTOM_SESSION_LIMIT=2000" >> "$GITHUB_ENV"
          echo "PHANTOM_RATE_LIMIT=1000000" >> "$GITHUB_ENV"
      - name: Install k6
        uses: grafana/setup-k6-action@v1
      - name: Verify Docker Compose availability
        run: docker compose version
      - name: Start observability stack
        env:
          TAG: sha-${{ github.sha }}
        run: |
          docker compose -f docker-compose.yml up -d prometheus grafana node-exporter
          timeout 60 bash -c 'until curl -sf http://localhost:9090/-/ready; do sleep 2; done'
          echo "Prometheus ready"
      - name: Start Phantom engine
        env:
          TAG: sha-${{ github.sha }}
        run: |
          docker compose -f docker-compose.yml up -d phantom
          timeout 60 bash -c 'until curl -sf http://localhost:8080/health; do sleep 2; done'
          echo "Phantom engine ready"
      - name: Run k6 navigate load
        run: k6 run --env PHANTOM_API_KEY="${PHANTOM_API_KEY}" --env PHANTOM_BASE_URL=http://localhost:8080 k6/navigate_load.js
      - name: Run k6 scene graph load
        run: k6 run --env PHANTOM_API_KEY="${PHANTOM_API_KEY}" --env PHANTOM_BASE_URL=http://localhost:8080 k6/scene_graph_load.js
      - name: Run k6 click load
        run: k6 run --env PHANTOM_API_KEY="${PHANTOM_API_KEY}" --env PHANTOM_BASE_URL=http://localhost:8080 k6/click_load.js
      - name: Run k6 mixed load
        run: k6 run --env PHANTOM_API_KEY="${PHANTOM_API_KEY}" --env PHANTOM_BASE_URL=http://localhost:8080 k6/mixed_load.js
      - name: Verify Prometheus metrics are populated
        run: |
          echo "Checking direct engine metrics..."
          curl -sf http://localhost:8080/metrics | grep sessions_created_total || echo "Warning: Metric not found"
          
          echo "Querying Prometheus (with retries)..."
          for i in {1..10}; do
            # Query sessions_created_total which persists even after sessions close
            response=$(curl -s -G 'http://localhost:9090/api/v1/query' --data-urlencode 'query=sessions_created_total')
            echo "Prometheus response: $response"
            
            result=$(echo "$response" | python3 -c "import json, sys; d=json.load(sys.stdin); print(sum(int(r['value'][1]) for r in d.get('data', {}).get('result', []))) if d.get('status') == 'success' else print(0)")
            if [ "$result" -gt 0 ]; then
              echo "Metrics confirmed! Total sessions created: $result"
              exit 0
            fi
            echo "Metrics not populated yet, waiting... ($i/10)"
            sleep 5
          done
          echo "Error: Metrics failed to populate in Prometheus after 50 seconds"
          exit 1
      - name: Print final metric snapshot
        if: always()
        run: curl -s http://localhost:8080/metrics | grep -E '^(sessions|phantom_sessions|js_|phantom_js|http_|phantom_http|dom_|phantom_dom|storage_|phantom_storage)'
      - name: Print compose logs on failure
        if: failure()
        run: docker compose -f docker-compose.yml logs --tail=200
      - name: Tear down compose stack
        if: always()
        run: docker compose -f docker-compose.yml down -v