build: parallelize Docker builds for ~50% speed improvement (#21)

deanq · web-flow · commit a38f50a42c60 · 2025-08-06T20:49:41.000-07:00
* build: parallelize Docker builds for ~50% speed improvement

Split sequential Docker builds into parallel GPU/CPU jobs:
- docker-pr → docker-pr-gpu + docker-pr-cpu (parallel)
- docker-main → docker-main-gpu + docker-main-cpu (parallel)
- docker-prod → docker-prod-gpu + docker-prod-cpu (parallel)

Benefits:
- GPU and CPU images build simultaneously instead of sequentially
- Reduces CI/CD Docker build time by approximately 50%
- Better GitHub Actions runner resource utilization
- Maintains all existing functionality and outputs

* build: fix redundant Docker builds during releases

Prevent docker-main jobs from running during releases by:
- Adding 'release' to needs dependencies
- Excluding when release_created is true

Now:
- Regular main pushes → docker-main jobs run (:main tags)
- Release pushes → only docker-prod jobs run (semantic tags)
- Eliminates redundant builds, saves CI/CD resources

* chore: update deps, claude, and make smoketest-macos

* build: fix Docker release tagging to generate exact version and latest tags

- Fix conditional logic: only run docker-prod jobs on actual releases
- Simplify tag patterns: generate only {{version}} and latest tags
- Results in exactly 4 tags per release:
  - runpod/tetra-rp:latest, runpod/tetra-rp:v1.2.3
  - runpod/tetra-rp-cpu:latest, runpod/tetra-rp-cpu:v1.2.3

* chore: cleanup
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -130,7 +130,7 @@ jobs:
           release-type: python
           token: ${{ secrets.GITHUB_TOKEN }}
 
-  docker-pr:
+  docker-pr-gpu:
     runs-on: ubuntu-latest
     needs: [test, lint, local-execution-test]
     if: github.event_name == 'pull_request'
@@ -177,6 +177,46 @@ jobs:
           cache-to: type=gha,mode=max
           load: true
 
+      - name: Test GPU Docker image with handler test suite
+        run: |
+          echo "Testing GPU image with handler test suite..."
+          docker run --rm tetra-rp:pr-test ./test-handler.sh
+
+  docker-pr-cpu:
+    runs-on: ubuntu-latest
+    needs: [test, lint, local-execution-test]
+    if: github.event_name == 'pull_request'
+    steps:
+      - name: Clear Space
+        run: |
+          rm -rf /usr/share/dotnet
+          rm -rf /opt/ghc
+          rm -rf "/usr/local/share/boost"
+          rm -rf "$AGENT_TOOLSDIRECTORY"
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          submodules: recursive
+          fetch-depth: 0
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Set up uv
+        uses: astral-sh/setup-uv@v4
+        with:
+          enable-cache: true
+
+      - name: Setup dependencies
+        run: |
+          uv sync
+          git submodule update --remote --merge
+          cp tetra-rp/src/tetra_rp/protos/remote_execution.py src/
+
       - name: Build and test CPU Docker image (PR)
         uses: docker/build-push-action@v6
         with:
@@ -189,20 +229,15 @@ jobs:
           cache-to: type=gha,mode=max
           load: true
 
-      - name: Test GPU Docker image with handler test suite
-        run: |
-          echo "Testing GPU image with handler test suite..."
-          docker run --rm tetra-rp:pr-test ./test-handler.sh
-
       - name: Test CPU Docker image with handler test suite
         run: |
           echo "Testing CPU image with handler test suite..."
           docker run --rm tetra-rp-cpu:pr-test ./test-handler.sh
 
-  docker-main:
+  docker-main-gpu:
     runs-on: ubuntu-latest
-    needs: [test, lint, local-execution-test]
-    if: github.ref == 'refs/heads/main' && github.event_name == 'push'
+    needs: [test, lint, local-execution-test, release]
+    if: github.ref == 'refs/heads/main' && github.event_name == 'push' && !needs.release.outputs.release_created
     steps:
       - name: Clear Space
         run: |
@@ -252,6 +287,48 @@ jobs:
           cache-from: type=gha
           cache-to: type=gha,mode=max
 
+  docker-main-cpu:
+    runs-on: ubuntu-latest
+    needs: [test, lint, local-execution-test, release]
+    if: github.ref == 'refs/heads/main' && github.event_name == 'push' && !needs.release.outputs.release_created
+    steps:
+      - name: Clear Space
+        run: |
+          rm -rf /usr/share/dotnet
+          rm -rf /opt/ghc
+          rm -rf "/usr/local/share/boost"
+          rm -rf "$AGENT_TOOLSDIRECTORY"
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          submodules: recursive
+          fetch-depth: 0
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Set up uv
+        uses: astral-sh/setup-uv@v4
+        with:
+          enable-cache: true
+
+      - name: Setup dependencies
+        run: |
+          uv sync
+          git submodule update --remote --merge
+          cp tetra-rp/src/tetra_rp/protos/remote_execution.py src/
+
       - name: Build and push CPU Docker image (main)
         uses: docker/build-push-action@v6
         with:
@@ -264,10 +341,10 @@ jobs:
           cache-to: type=gha,mode=max
 
 
-  docker-prod:
+  docker-prod-gpu:
     runs-on: ubuntu-latest
     needs: [release]
-    if: needs.release.outputs.release_created || (github.ref == 'refs/heads/main' && needs.release.outputs.tag_name)
+    if: needs.release.outputs.release_created
     steps:
       - name: Clear Space
         run: |
@@ -301,24 +378,7 @@ jobs:
         with:
           images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
           tags: |
-            type=ref,event=branch
-            type=ref,event=pr
-            type=semver,pattern={{version}}
-            type=semver,pattern={{major}}.{{minor}}
-            type=semver,pattern={{major}}
-            type=raw,value=latest,enable={{is_default_branch}}
-
-      - name: Extract CPU metadata
-        id: meta-cpu
-        uses: docker/metadata-action@v5
-        with:
-          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}-cpu
-          tags: |
-            type=ref,event=branch
-            type=ref,event=pr
             type=semver,pattern={{version}}
-            type=semver,pattern={{major}}.{{minor}}
-            type=semver,pattern={{major}}
             type=raw,value=latest,enable={{is_default_branch}}
 
       - name: Set up uv
@@ -344,6 +404,57 @@ jobs:
           cache-from: type=gha
           cache-to: type=gha,mode=max
 
+  docker-prod-cpu:
+    runs-on: ubuntu-latest
+    needs: [release]
+    if: needs.release.outputs.release_created
+    steps:
+      - name: Clear Space
+        run: |
+          rm -rf /usr/share/dotnet
+          rm -rf /opt/ghc
+          rm -rf "/usr/local/share/boost"
+          rm -rf "$AGENT_TOOLSDIRECTORY"
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          submodules: recursive
+          fetch-depth: 0
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Extract CPU metadata
+        id: meta-cpu
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}-cpu
+          tags: |
+            type=semver,pattern={{version}}
+            type=raw,value=latest,enable={{is_default_branch}}
+
+      - name: Set up uv
+        uses: astral-sh/setup-uv@v4
+        with:
+          enable-cache: true
+
+      - name: Setup dependencies
+        run: |
+          uv sync
+          git submodule update --remote --merge
+          cp tetra-rp/src/tetra_rp/protos/remote_execution.py src/
+
       - name: Build and push CPU Docker image (prod)
         uses: docker/build-push-action@v6
         with:
diff --git a/CLAUDE.md b/CLAUDE.md
@@ -255,4 +255,9 @@ Configure these in GitHub repository settings:
 
 ## Development Best Practices
 
-- Always run `make quality-check` before committing changes
+- Always run `make quality-check` before committing changes
+
+## Project Memories
+
+### Docker Guidelines
+- Docker container should never refer to src/
diff --git a/Makefile b/Makefile
@@ -70,6 +70,18 @@ test-fast: # Run tests with fast-fail mode
 test-handler: # Test handler locally with all test_*.json files
 	./test-handler.sh
 
+# Smoke Tests (local on Mac OS)
+
+smoketest-macos-build: setup # Build CPU-only Mac OS Docker image (macos/arm64)
+	docker buildx build \
+	--platform linux/arm64 \
+	-f Dockerfile-cpu \
+	-t $(FULL_IMAGE_CPU)-mac \
+	. --load
+
+smoketest-macos: smoketest-macos-build # Test CPU Docker image locally
+	docker run --rm $(FULL_IMAGE_CPU)-mac ./test-handler.sh
+
 # Linting commands
 lint: # Check code with ruff
 	uv run ruff check .
diff --git a/uv.lock b/uv.lock
