fix(ci): simplify workflow to matrix strategy for Gitea compat
Some checks failed
Build and Push Images / determine-version (push) Successful in 5s
Build and Push Images / build (Dockerfile.ray-worker-nvidia, nvidia) (push) Failing after 23m46s
Build and Push Images / build (Dockerfile.ray-worker-intel, intel) (push) Failing after 23m50s
Build and Push Images / build (Dockerfile.ray-worker-strixhalo, strixhalo) (push) Failing after 31s
Build and Push Images / build (Dockerfile.ray-worker-rdna2, rdna2) (push) Failing after 43s
Build and Push Images / Release (push) Has been skipped
Build and Push Images / Notify (push) Successful in 2s

- Replace 4 separate build jobs with single matrix build job
- Eliminates complex dependency graph causing 'must contain one job
  without dependencies' parse error in Gitea act_runner
- All if: conditions now use single-line strings (no multi-line |)
- workflow_dispatch image filter moved to step-level check
- Add stale buildx builder cleanup step before each build
- Simplify release/notify to depend on single 'build' job
This commit is contained in:
2026-02-06 15:41:19 -05:00
parent 3bc0b848de
commit 98c3ef284f

View File

@@ -88,19 +88,55 @@ jobs:
echo "should_release=true" >> $GITHUB_OUTPUT
echo "📦 Version: $LATEST → $NEW_VERSION ($BUMP bump)"
build-nvidia:
# 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]
# Skip if commit message contains [skip images] or [ray-serve only]
if: |
!contains(github.event.head_commit.message, '[skip images]') &&
!contains(github.event.head_commit.message, '[ray-serve only]') &&
(github.event_name != 'workflow_dispatch' || github.event.inputs.image == 'all' || github.event.inputs.image == 'nvidia')
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: |
@@ -108,17 +144,15 @@ jobs:
http = true
insecure = true
# Login to Docker Hub to avoid pull rate limits
- name: Login to Docker Hub
if: vars.DOCKERHUB_USERNAME != ''
if: steps.check.outputs.skip != 'true' && vars.DOCKERHUB_USERNAME != ''
uses: docker/login-action@v3
with:
username: ${{ vars.DOCKERHUB_USERNAME }}
password: ${{ secrets.DOCKERHUB_TOKEN }}
# Configure Gitea registry auth (create config.json directly for HTTP registry)
- name: Configure Gitea Registry Auth
if: github.event_name != 'pull_request'
if: steps.check.outputs.skip != 'true' && github.event_name != 'pull_request'
run: |
AUTH=$(echo -n "${{ secrets.REGISTRY_USER }}:${{ secrets.REGISTRY_TOKEN }}" | base64 -w0)
mkdir -p ~/.docker
@@ -134,219 +168,32 @@ jobs:
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-nvidia
images: ${{ env.REGISTRY }}/ray-worker-${{ matrix.name }}
tags: |
type=raw,value=${{ needs.determine-version.outputs.version }}
type=raw,value=latest,enable={{is_default_branch}}
- name: Build and push
if: steps.check.outputs.skip != 'true'
uses: docker/build-push-action@v5
with:
context: .
file: dockerfiles/Dockerfile.ray-worker-nvidia
file: dockerfiles/${{ 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-nvidia:buildcache
cache-to: type=registry,ref=${{ env.REGISTRY }}/ray-worker-nvidia:buildcache,mode=max,image-manifest=true,compression=zstd
build-rdna2:
needs: [determine-version]
if: |
!contains(github.event.head_commit.message, '[skip images]') &&
!contains(github.event.head_commit.message, '[ray-serve only]') &&
(github.event_name != 'workflow_dispatch' || github.event.inputs.image == 'all' || github.event.inputs.image == 'rdna2')
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Docker Buildx
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: vars.DOCKERHUB_USERNAME != ''
uses: docker/login-action@v3
with:
username: ${{ vars.DOCKERHUB_USERNAME }}
password: ${{ secrets.DOCKERHUB_TOKEN }}
# Configure Gitea registry auth (create config.json directly for HTTP registry)
- name: Configure Gitea Registry Auth
if: github.event_name != 'pull_request'
run: |
AUTH=$(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
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY }}/ray-worker-rdna2
tags: |
type=raw,value=${{ needs.determine-version.outputs.version }}
type=raw,value=latest,enable={{is_default_branch}}
- name: Build and push
uses: docker/build-push-action@v5
with:
context: .
file: dockerfiles/Dockerfile.ray-worker-rdna2
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-rdna2:buildcache
cache-to: type=registry,ref=${{ env.REGISTRY }}/ray-worker-rdna2:buildcache,mode=max,image-manifest=true,compression=zstd
build-strixhalo:
needs: [determine-version]
if: |
!contains(github.event.head_commit.message, '[skip images]') &&
!contains(github.event.head_commit.message, '[ray-serve only]') &&
(github.event_name != 'workflow_dispatch' || github.event.inputs.image == 'all' || github.event.inputs.image == 'strixhalo')
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Docker Buildx
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: vars.DOCKERHUB_USERNAME != ''
uses: docker/login-action@v3
with:
username: ${{ vars.DOCKERHUB_USERNAME }}
password: ${{ secrets.DOCKERHUB_TOKEN }}
# Configure Gitea registry auth (create config.json directly for HTTP registry)
- name: Configure Gitea Registry Auth
if: github.event_name != 'pull_request'
run: |
AUTH=$(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
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY }}/ray-worker-strixhalo
tags: |
type=raw,value=${{ needs.determine-version.outputs.version }}
type=raw,value=latest,enable={{is_default_branch}}
- name: Build and push
uses: docker/build-push-action@v5
with:
context: .
file: dockerfiles/Dockerfile.ray-worker-strixhalo
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-strixhalo:buildcache
cache-to: type=registry,ref=${{ env.REGISTRY }}/ray-worker-strixhalo:buildcache,mode=max,image-manifest=true,compression=zstd
build-intel:
needs: [determine-version]
if: |
!contains(github.event.head_commit.message, '[skip images]') &&
!contains(github.event.head_commit.message, '[ray-serve only]') &&
(github.event_name != 'workflow_dispatch' || github.event.inputs.image == 'all' || github.event.inputs.image == 'intel')
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Docker Buildx
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: vars.DOCKERHUB_USERNAME != ''
uses: docker/login-action@v3
with:
username: ${{ vars.DOCKERHUB_USERNAME }}
password: ${{ secrets.DOCKERHUB_TOKEN }}
# Configure Gitea registry auth (create config.json directly for HTTP registry)
- name: Configure Gitea Registry Auth
if: github.event_name != 'pull_request'
run: |
AUTH=$(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
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY }}/ray-worker-intel
tags: |
type=raw,value=${{ needs.determine-version.outputs.version }}
type=raw,value=latest,enable={{is_default_branch}}
- name: Build and push
uses: docker/build-push-action@v5
with:
context: .
file: dockerfiles/Dockerfile.ray-worker-intel
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-intel:buildcache
cache-to: type=registry,ref=${{ env.REGISTRY }}/ray-worker-intel:buildcache,mode=max,image-manifest=true,compression=zstd
cache-from: type=registry,ref=${{ 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-nvidia, build-rdna2, build-strixhalo, build-intel]
if: needs.determine-version.outputs.should_release == 'true' && github.ref == 'refs/heads/main' && github.event_name == 'push'
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
@@ -366,16 +213,11 @@ jobs:
notify:
name: Notify
runs-on: ubuntu-latest
# Don't depend on release - notify should run even if release never starts
needs: [determine-version, build-nvidia, build-rdna2, build-strixhalo, build-intel]
needs: [determine-version, build]
if: always()
steps:
- name: Notify on success
if: |
(needs.build-nvidia.result == 'success' || needs.build-nvidia.result == 'skipped') &&
(needs.build-rdna2.result == 'success' || needs.build-rdna2.result == 'skipped') &&
(needs.build-strixhalo.result == 'success' || needs.build-strixhalo.result == 'skipped') &&
(needs.build-intel.result == 'success' || needs.build-intel.result == 'skipped')
if: "needs.build.result == 'success' || needs.build.result == 'skipped'"
run: |
curl -s \
-H "Title: ✅ Images Built: ${{ gitea.repository }}" \
@@ -384,18 +226,11 @@ jobs:
-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 }})
nvidia: ${{ needs.build-nvidia.result }}
rdna2: ${{ needs.build-rdna2.result }}
strixhalo: ${{ needs.build-strixhalo.result }}
intel: ${{ needs.build-intel.result }}" \
Build: ${{ needs.build.result }}" \
${{ env.NTFY_URL }}/gitea-ci
- name: Notify on failure or cancellation
if: |
needs.build-nvidia.result == 'failure' || needs.build-nvidia.result == 'cancelled' ||
needs.build-rdna2.result == 'failure' || needs.build-rdna2.result == 'cancelled' ||
needs.build-strixhalo.result == 'failure' || needs.build-strixhalo.result == 'cancelled' ||
needs.build-intel.result == 'failure' || needs.build-intel.result == 'cancelled'
if: "needs.build.result == 'failure' || needs.build.result == 'cancelled'"
run: |
curl -s \
-H "Title: ❌ Image Build Failed: ${{ gitea.repository }}" \
@@ -404,8 +239,5 @@ jobs:
-H "Click: ${{ gitea.server_url }}/${{ gitea.repository }}/actions/runs/${{ gitea.run_id }}" \
-d "Branch: ${{ gitea.ref_name }}
Version: ${{ needs.determine-version.outputs.version }}
nvidia: ${{ needs.build-nvidia.result }}
rdna2: ${{ needs.build-rdna2.result }}
strixhalo: ${{ needs.build-strixhalo.result }}
intel: ${{ needs.build-intel.result }}" \
Build: ${{ needs.build.result }}" \
${{ env.NTFY_URL }}/gitea-ci