kuberay-images/.gitea/workflows/build-push.yaml

name: Build and Push Images

on:
  push:
    branches:
      - main
    tags:
      - 'v*'
    paths:
      - 'dockerfiles/**'
      - 'amdsmi-shim/**'
      - '.gitea/workflows/build-push.yaml'
  pull_request:
    branches:
      - main
    paths:
      - 'dockerfiles/**'
      - 'amdsmi-shim/**'
      - '.gitea/workflows/build-push.yaml'
  workflow_dispatch:
    inputs:
      image:
        description: 'Image to build (all, nvidia, rdna2, strixhalo, intel)'
        required: false
        default: 'all'

env:
  # Use internal HTTP endpoint (avoids Cloudflare 100MB limit and HTTPS issues)
  REGISTRY: gitea-http.gitea.svc.cluster.local:3000/daviestechlabs
  REGISTRY_HOST: gitea-http.gitea.svc.cluster.local:3000
  NTFY_URL: http://ntfy.observability.svc.cluster.local:80

jobs:
  # Determine semantic version BEFORE building images
  determine-version:
    runs-on: ubuntu-latest
    outputs:
      version: ${{ steps.version.outputs.version }}
      bump: ${{ steps.version.outputs.bump }}
      should_release: ${{ steps.version.outputs.should_release }}
    steps:
      - name: Checkout
        uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - name: Calculate semantic version
        id: version
        run: |
          # Skip version calculation for PRs and tag pushes
          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
            echo "version=pr-${{ github.event.number }}" >> $GITHUB_OUTPUT
            echo "bump=none" >> $GITHUB_OUTPUT
            echo "should_release=false" >> $GITHUB_OUTPUT
            exit 0
          fi
          
          # For tag pushes, use the tag directly
          if [[ "${{ github.ref }}" == refs/tags/v* ]]; then
            VERSION="${{ github.ref_name }}"
            echo "version=$VERSION" >> $GITHUB_OUTPUT
            echo "bump=tag" >> $GITHUB_OUTPUT
            echo "should_release=false" >> $GITHUB_OUTPUT
            exit 0
          fi
          
          # Get latest tag or default to v0.0.0
          LATEST=$(git describe --tags --abbrev=0 2>/dev/null || echo "v0.0.0")
          VERSION=${LATEST#v}
          IFS='.' read -r MAJOR MINOR PATCH <<< "$VERSION"
          
          # Check commit message for version bump keywords
          MSG="${{ github.event.head_commit.message }}"
          if echo "$MSG" | grep -qiE "^major:|BREAKING CHANGE"; then
            MAJOR=$((MAJOR + 1)); MINOR=0; PATCH=0
            BUMP="major"
          elif echo "$MSG" | grep -qiE "^(minor:|feat:|feature:)"; then
            MINOR=$((MINOR + 1)); PATCH=0
            BUMP="minor"
          else
            PATCH=$((PATCH + 1))
            BUMP="patch"
          fi
          
          NEW_VERSION="v${MAJOR}.${MINOR}.${PATCH}"
          echo "version=$NEW_VERSION" >> $GITHUB_OUTPUT
          echo "bump=$BUMP" >> $GITHUB_OUTPUT
          echo "should_release=true" >> $GITHUB_OUTPUT
          echo "📦 Version: $LATEST → $NEW_VERSION ($BUMP bump)"

  # Build all GPU worker images using a matrix strategy
  # Runs sequentially on single runner, but one job definition = simpler dependency graph
  build:
    needs: [determine-version]
    if: "!contains(github.event.head_commit.message, '[skip images]') && !contains(github.event.head_commit.message, '[ray-serve only]')"
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
      matrix:
        include:
          - name: nvidia
            dockerfile: Dockerfile.ray-worker-nvidia
          - name: rdna2
            dockerfile: Dockerfile.ray-worker-rdna2
          - name: strixhalo
            dockerfile: Dockerfile.ray-worker-strixhalo
          - name: intel
            dockerfile: Dockerfile.ray-worker-intel
    steps:
      # For workflow_dispatch, allow building a single image
      - name: Check if should build this variant
        id: check
        run: |
          if [[ "${{ github.event_name }}" == "workflow_dispatch" ]]; then
            INPUT="${{ github.event.inputs.image }}"
            if [[ -n "$INPUT" && "$INPUT" != "all" && "$INPUT" != "${{ matrix.name }}" ]]; then
              echo "skip=true" >> $GITHUB_OUTPUT
              echo "⏭ Skipping ${{ matrix.name }} (dispatch requested: $INPUT)"
              exit 0
            fi
          fi
          echo "skip=false" >> $GITHUB_OUTPUT
          echo "🔨 Building ${{ matrix.name }}"

      - name: Checkout
        if: steps.check.outputs.skip != 'true'
        uses: actions/checkout@v4

      # Clean stale buildx builders from persistent Docker cache (prevents GPU capability errors)
      - name: Clean stale buildx builders
        if: steps.check.outputs.skip != 'true'
        run: |
          for b in $(docker buildx ls --format '{{.Name}}' 2>/dev/null | grep -v default | grep -v '\*' | sort -u); do
            docker buildx rm "$b" --force 2>/dev/null || true
          done
          docker container prune --force 2>/dev/null || true

      - name: Set up Docker Buildx
        if: steps.check.outputs.skip != 'true'
        uses: docker/setup-buildx-action@v3
        with:
          buildkitd-config-inline: |
            [registry."gitea-http.gitea.svc.cluster.local:3000"]
              http = true
              insecure = true

      - name: Login to Docker Hub
        if: steps.check.outputs.skip != 'true' && vars.DOCKERHUB_USERNAME != ''
        uses: docker/login-action@v3
        with:
          username: ${{ vars.DOCKERHUB_USERNAME }}
          password: ${{ secrets.DOCKERHUB_TOKEN }}

      - name: Configure Gitea Registry Auth
        if: steps.check.outputs.skip != 'true' && github.event_name != 'pull_request'
        run: |
          AUTH=$(echo -n "${{ secrets.REGISTRY_USER }}:${{ secrets.REGISTRY_TOKEN }}" | base64 -w0)
          mkdir -p ~/.docker
          cat > ~/.docker/config.json << EOF
          {
            "auths": {
              "${{ env.REGISTRY_HOST }}": {
                "auth": "$AUTH"
              }
            }
          }
          EOF
          echo "Auth configured for ${{ env.REGISTRY_HOST }}"

      - name: Extract metadata
        if: steps.check.outputs.skip != 'true'
        id: meta
        uses: docker/metadata-action@v5
        with:
          images: ${{ env.REGISTRY }}/ray-worker-${{ matrix.name }}
          tags: |
            type=raw,value=${{ needs.determine-version.outputs.version }}
            type=raw,value=latest,enable={{is_default_branch}}

      - name: Build and push
        if: steps.check.outputs.skip != 'true'
        uses: docker/build-push-action@v5
        with:
          context: .
          file: dockerfiles/${{ matrix.dockerfile }}
          push: ${{ github.event_name != 'pull_request' }}
          tags: ${{ steps.meta.outputs.tags }}
          labels: ${{ steps.meta.outputs.labels }}
          cache-from: type=registry,ref=${{ env.REGISTRY }}/ray-worker-${{ matrix.name }}:buildcache
          cache-to: type=registry,ref=${{ env.REGISTRY }}/ray-worker-${{ matrix.name }}:buildcache,mode=max,image-manifest=true,compression=zstd

  release:
    name: Release
    runs-on: ubuntu-latest
    needs: [determine-version, build]
    if: "needs.determine-version.outputs.should_release == 'true' && github.ref == 'refs/heads/main' && github.event_name == 'push'"
    steps:
      - name: Checkout
        uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - name: Create and push tag
        run: |
          VERSION="${{ needs.determine-version.outputs.version }}"
          BUMP="${{ needs.determine-version.outputs.bump }}"
          echo "📦 Creating release tag: $VERSION ($BUMP bump)"
          git config user.name "gitea-actions[bot]"
          git config user.email "actions@git.daviestechlabs.io"
          git tag -a "$VERSION" -m "Release $VERSION ($BUMP)"
          git push origin "$VERSION"

  notify:
    name: Notify
    runs-on: ubuntu-latest
    needs: [determine-version, build]
    if: always()
    steps:
      - name: Notify on success
        if: "needs.build.result == 'success' || needs.build.result == 'skipped'"
        run: |
          curl -s \
            -H "Title: ✅ Images Built: ${{ gitea.repository }}" \
            -H "Priority: default" \
            -H "Tags: white_check_mark,docker" \
            -H "Click: ${{ gitea.server_url }}/${{ gitea.repository }}/actions/runs/${{ gitea.run_id }}" \
            -d "Branch: ${{ gitea.ref_name }}
          Version: ${{ needs.determine-version.outputs.version }} (${{ needs.determine-version.outputs.bump }})
          Build: ${{ needs.build.result }}" \
            ${{ env.NTFY_URL }}/gitea-ci

      - name: Notify on failure or cancellation
        if: "needs.build.result == 'failure' || needs.build.result == 'cancelled'"
        run: |
          curl -s \
            -H "Title: ❌ Image Build Failed: ${{ gitea.repository }}" \
            -H "Priority: high" \
            -H "Tags: x,docker" \
            -H "Click: ${{ gitea.server_url }}/${{ gitea.repository }}/actions/runs/${{ gitea.run_id }}" \
            -d "Branch: ${{ gitea.ref_name }}
          Version: ${{ needs.determine-version.outputs.version }}
          Build: ${{ needs.build.result }}" \
            ${{ env.NTFY_URL }}/gitea-ci