ci(release): native arm64 runners + split-build manifest pattern; drop OTel variant

Root cause of v3.10.0/v3.11.0 release hangs (cancelled at 6h timeout):
the docker-images (full) job built linux/arm64 under QEMU emulation while
ENABLE_FULL_SKILLS=true triggered pip install of pandas + lxml, whose C
extensions have to compile from source on musl arm64 (no prebuilt wheels).
QEMU compile time for pandas alone exceeded the 6h job ceiling.

Changes:
- docker-images split into docker-build (per variant × platform) +
  docker-merge (manifest list per variant). amd64 runs on ubuntu-latest;
  arm64 runs on ubuntu-24.04-arm — native, no QEMU. Pandas/lxml compile
  natively in ~10-15 min instead of >6h.
- docker-web likewise split into docker-web-build + docker-web-merge.
- Drop the otel variant — niche, users build with --build-arg
  ENABLE_OTEL=true. Tailscale was never pre-built (build tag only).
- Add timeout-minutes per job (5–90 depending on workload) so failures
  surface in minutes rather than hours.
- provenance: false to skip buildx attestation overhead (~5-10% faster).
- cache scope per (variant, platform) for higher hit rate on rerun.

CLAUDE.md docker variants table updated to reflect dropped OTel.
This commit is contained in:
viettranx
2026-04-25 07:32:36 +07:00
parent c4bf2ed460
commit ef06bd445f
2 changed files with 186 additions and 53 deletions
+184 -52
View File
@@ -22,6 +22,7 @@ env:
jobs:
release:
runs-on: ubuntu-latest
timeout-minutes: 5
outputs:
version: ${{ steps.version.outputs.version }}
released: ${{ steps.version.outputs.version != '' }}
@@ -33,7 +34,6 @@ jobs:
- name: Resolve version from tag
id: version
run: |
# Priority: workflow_dispatch input > pushed tag > latest tag on HEAD
TAG="${{ inputs.tag || github.ref_name }}"
if [[ -z "$TAG" || "$TAG" == "main" || "$TAG" == "dev" ]]; then
TAG=$(git describe --tags --abbrev=0 HEAD 2>/dev/null || echo "")
@@ -65,7 +65,9 @@ jobs:
needs: release
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
include:
- goos: linux
@@ -119,6 +121,7 @@ jobs:
needs: [release, build-binaries]
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- name: Download all binary assets
env:
@@ -146,46 +149,106 @@ jobs:
--repo "${{ github.repository }}" \
--clobber
# Build and push Docker images to GHCR + Docker Hub
docker-images:
# Build Docker images per (variant × platform) on native runners (no QEMU).
# Each job pushes a digest-only image; the merge job combines them into a
# multi-arch manifest and pushes the final tags to both GHCR and Docker Hub.
# Variants: base (backend only), latest (+ web UI + Python), full (+ all skills).
# OTel and Tailscale variants are not built — users build from source if needed.
docker-build:
needs: release
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 90
strategy:
fail-fast: false
matrix:
variant: [base, latest, full]
platform: [linux/amd64, linux/arm64]
include:
# Base: backend API-only, no web UI, no runtimes
- variant: base
suffix: "-base"
enable_otel: "false"
enable_embedui: "false"
enable_python: "false"
enable_full_skills: "false"
# Latest: backend + embedded web UI + Python
- variant: latest
suffix: ""
enable_otel: "false"
enable_embedui: "true"
enable_python: "true"
enable_full_skills: "false"
# Full: all runtimes + skills pre-installed
- variant: full
suffix: "-full"
enable_otel: "false"
enable_embedui: "true"
enable_python: "true"
enable_full_skills: "true"
# OTel: latest + OpenTelemetry tracing
- variant: otel
suffix: "-otel"
enable_otel: "true"
enable_embedui: "true"
enable_python: "true"
enable_full_skills: "false"
- platform: linux/amd64
runner: ubuntu-latest
arch: amd64
- platform: linux/arm64
runner: ubuntu-24.04-arm
arch: arm64
runs-on: ${{ matrix.runner }}
steps:
- uses: actions/checkout@v4
- uses: docker/setup-qemu-action@v3
- uses: docker/setup-buildx-action@v3
- name: Log in to GHCR
uses: docker/login-action@v3
with:
registry: ghcr.io
username: ${{ github.actor }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Build and push by digest
id: build
uses: docker/build-push-action@v6
with:
context: .
platforms: ${{ matrix.platform }}
outputs: type=image,name=${{ env.GHCR_IMAGE }},push-by-digest=true,name-canonical=true,push=true
build-args: |
ENABLE_OTEL=false
ENABLE_EMBEDUI=${{ matrix.enable_embedui }}
ENABLE_PYTHON=${{ matrix.enable_python }}
ENABLE_FULL_SKILLS=${{ matrix.enable_full_skills }}
VERSION=v${{ needs.release.outputs.version }}
cache-from: type=gha,scope=${{ matrix.variant }}-${{ matrix.arch }}
cache-to: type=gha,mode=max,scope=${{ matrix.variant }}-${{ matrix.arch }}
provenance: false
- name: Export digest
run: |
mkdir -p /tmp/digests
digest="${{ steps.build.outputs.digest }}"
touch "/tmp/digests/${digest#sha256:}"
- uses: actions/upload-artifact@v4
with:
name: digests-${{ matrix.variant }}-${{ matrix.arch }}
path: /tmp/digests/*
if-no-files-found: error
retention-days: 1
# Combine per-platform digests into multi-arch manifests and push final tags
# to both GHCR and Docker Hub. One job per variant runs in parallel.
docker-merge:
needs: [release, docker-build]
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 15
strategy:
fail-fast: false
matrix:
include:
- variant: base
suffix: "-base"
- variant: latest
suffix: ""
- variant: full
suffix: "-full"
steps:
- name: Download digests
uses: actions/download-artifact@v4
with:
path: /tmp/digests
pattern: digests-${{ matrix.variant }}-*
merge-multiple: true
- uses: docker/setup-buildx-action@v3
@@ -214,32 +277,92 @@ jobs:
type=raw,value=latest,enable=${{ matrix.suffix == '' }},suffix=
type=raw,value=${{ matrix.variant }},enable=${{ matrix.suffix != '' }}
- name: Build and push
uses: docker/build-push-action@v6
with:
context: .
platforms: linux/amd64,linux/arm64
push: true
tags: ${{ steps.meta.outputs.tags }}
labels: ${{ steps.meta.outputs.labels }}
build-args: |
ENABLE_OTEL=${{ matrix.enable_otel }}
ENABLE_EMBEDUI=${{ matrix.enable_embedui }}
ENABLE_PYTHON=${{ matrix.enable_python }}
ENABLE_FULL_SKILLS=${{ matrix.enable_full_skills }}
VERSION=v${{ needs.release.outputs.version }}
cache-from: type=gha,scope=${{ matrix.variant }}
cache-to: type=gha,mode=max,scope=${{ matrix.variant }}
- name: Create manifest list and push
working-directory: /tmp/digests
run: |
# Compose -t flags from metadata-action tags (newline-separated)
TAG_ARGS=()
while IFS= read -r tag; do
[ -n "$tag" ] && TAG_ARGS+=("-t" "$tag")
done <<< "${{ steps.meta.outputs.tags }}"
# Compose source digest refs (pushed to GHCR by docker-build)
DIGEST_ARGS=()
for f in *; do
DIGEST_ARGS+=("${{ env.GHCR_IMAGE }}@sha256:$f")
done
set -x
docker buildx imagetools create "${TAG_ARGS[@]}" "${DIGEST_ARGS[@]}"
# Build and push web UI Docker image
docker-web:
- name: Inspect manifest
run: |
docker buildx imagetools inspect \
"${{ env.GHCR_IMAGE }}:v${{ needs.release.outputs.version }}${{ matrix.suffix }}"
# Build web UI Docker image per platform on native runners.
docker-web-build:
needs: release
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
platform: [linux/amd64, linux/arm64]
include:
- platform: linux/amd64
runner: ubuntu-latest
arch: amd64
- platform: linux/arm64
runner: ubuntu-24.04-arm
arch: arm64
runs-on: ${{ matrix.runner }}
steps:
- uses: actions/checkout@v4
- uses: docker/setup-qemu-action@v3
- uses: docker/setup-buildx-action@v3
- name: Log in to GHCR
uses: docker/login-action@v3
with:
registry: ghcr.io
username: ${{ github.actor }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Build and push by digest
id: build
uses: docker/build-push-action@v6
with:
context: ui/web
platforms: ${{ matrix.platform }}
outputs: type=image,name=${{ env.GHCR_IMAGE }}-web,push-by-digest=true,name-canonical=true,push=true
cache-from: type=gha,scope=web-${{ matrix.arch }}
cache-to: type=gha,mode=max,scope=web-${{ matrix.arch }}
provenance: false
- name: Export digest
run: |
mkdir -p /tmp/digests
digest="${{ steps.build.outputs.digest }}"
touch "/tmp/digests/${digest#sha256:}"
- uses: actions/upload-artifact@v4
with:
name: digests-web-${{ matrix.arch }}
path: /tmp/digests/*
if-no-files-found: error
retention-days: 1
docker-web-merge:
needs: [release, docker-web-build]
if: needs.release.outputs.released == 'true'
runs-on: ubuntu-latest
timeout-minutes: 15
steps:
- name: Download digests
uses: actions/download-artifact@v4
with:
path: /tmp/digests
pattern: digests-web-*
merge-multiple: true
- uses: docker/setup-buildx-action@v3
@@ -267,22 +390,31 @@ jobs:
type=raw,value=v${{ needs.release.outputs.version }}
type=raw,value=latest
- name: Build and push
uses: docker/build-push-action@v6
with:
context: ui/web
platforms: linux/amd64,linux/arm64
push: true
tags: ${{ steps.meta.outputs.tags }}
labels: ${{ steps.meta.outputs.labels }}
cache-from: type=gha,scope=web
cache-to: type=gha,mode=max,scope=web
- name: Create manifest list and push
working-directory: /tmp/digests
run: |
TAG_ARGS=()
while IFS= read -r tag; do
[ -n "$tag" ] && TAG_ARGS+=("-t" "$tag")
done <<< "${{ steps.meta.outputs.tags }}"
DIGEST_ARGS=()
for f in *; do
DIGEST_ARGS+=("${{ env.GHCR_IMAGE }}-web@sha256:$f")
done
set -x
docker buildx imagetools create "${TAG_ARGS[@]}" "${DIGEST_ARGS[@]}"
- name: Inspect manifest
run: |
docker buildx imagetools inspect \
"${{ env.GHCR_IMAGE }}-web:v${{ needs.release.outputs.version }}"
# Notify Discord on new release (runs even if docker jobs fail)
notify-discord:
needs: [release, build-binaries, docker-images, docker-web]
needs: [release, build-binaries, docker-merge, docker-web-merge]
if: always() && needs.release.outputs.released == 'true' && !cancelled()
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- name: Send Discord notification
env: