fix: Deploy阶段全面修复 - E2E服务器错位/Worker venv统一/内网Registry登录/生产构建缓存

.gitea/workflows/deploy.yml

@@ -166,7 +166,7 @@ jobs:
           echo "ERROR: No SSH key available"
           exit 1
         fi
-        echo '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' | base64 -d | ssh -i "$key_path" "$staging_user@$staging_host" "GITHUB_SHA='${GITHUB_SHA}' sh"
+        echo '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' | base64 -d | ssh -i "$key_path" "$staging_user@$staging_host" "GITHUB_SHA='${GITHUB_SHA}' sh"
 
     - name: Post-deploy smoke test
       shell: sh
@@ -261,35 +261,48 @@ jobs:
     needs: deploy-staging
 
     steps:
-    - name: Checkout code
+    - name: Install SSH client
+      shell: sh
+      run: |
+        set -eu
+        apt-get update -qq && apt-get install -y -qq openssh-client >/dev/null 2>&1
+        echo "openssh-client installed"
+
+    - name: Run Playwright E2E on staging server
       shell: sh
       env:
-        GITHUB_TOKEN: ${{ github.token }}
+        STAGING_SSH_HOST: ${{ secrets.STAGING_SSH_HOST }}
+        STAGING_SSH_USER: ${{ secrets.STAGING_SSH_USER }}
+        STAGING_SSH_KEY: ${{ secrets.STAGING_SSH_KEY }}
       run: |
         set -eu
-        python3 - <<'PY'
-        import io, os, tarfile, urllib.request
-        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
-        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
-        with urllib.request.urlopen(request, timeout=120) as response:
-            archive = response.read()
-        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
-            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
-            for member in tar.getmembers():
-                name = member.name
-                if name == root_prefix[:-1]:
-                    continue
-                if name.startswith(root_prefix):
-                    member.name = name[len(root_prefix):]
-                if member.name:
-                    tar.extract(member, '.')
-        PY
+        staging_host="${STAGING_SSH_HOST:-47.98.113.167}"
+        staging_user="${STAGING_SSH_USER:-root}"
+        if [ -f /root/.ssh/xiaoxia_runtime_builder ]; then
+          key_path="/root/.ssh/xiaoxia_runtime_builder"
+        elif [ -n "${STAGING_SSH_KEY:-}" ]; then
+          key_path="$HOME/.ssh/id_ed25519"
+          printf '%s\n' "$STAGING_SSH_KEY" > "$key_path"
+          chmod 600 "$key_path"
+        else
+          echo "ERROR: No SSH key available"
+          exit 1
+        fi
+        ssh-keyscan -H "$staging_host" >> ~/.ssh/known_hosts
 
-    - name: Run Playwright E2E against staging
-      shell: sh
-      run: |
-        set -eu
-        docker run --rm           -e E2E_BASE_URL=http://127.0.0.1:3001           -e E2E_API_BASE=http://127.0.0.1:8000/api/v1           -e E2E_BROWSER_CHANNEL=chromium           -v "$PWD:/workspace"           -w /workspace/apps/web           --network host           mcr.microsoft.com/playwright:v1.45.0-jammy           sh -lc 'npm ci && npx playwright test --reporter=line --project=chromium'
+        # 在业务服务器上跑 Playwright E2E（用 host 网络访问 staging 3001/8000 端口）
+        ssh -i "$key_path" "$staging_user@$staging_host" '
+          cd /var/lib/xiaoxia-saas-staging/repo
+          docker run --rm \
+            -e E2E_BASE_URL=http://127.0.0.1:3001 \
+            -e E2E_API_BASE=http://127.0.0.1:8000/api/v1 \
+            -e E2E_BROWSER_CHANNEL=chromium \
+            -v "$PWD:/workspace" \
+            -w /workspace/apps/web \
+            --network host \
+            mcr.microsoft.com/playwright:v1.45.0-jammy \
+            sh -lc "npm ci && npx playwright test --reporter=line --project=chromium"
+        '
 
   build-production-runtime-images:
     name: Build Production Runtime Images
@@ -323,10 +336,12 @@ jobs:
 
     - name: Build runtime image artifact
       shell: sh
+      env:
+        REGISTRY_TOKEN: ${{ secrets.REGISTRY_TOKEN }}
       run: |
         set -eu
         chmod +x scripts/build_release_images.sh
-        scripts/build_release_images.sh "${GITHUB_REF_NAME}"
+        REGISTRY_TOKEN="${REGISTRY_TOKEN}" scripts/build_release_images.sh "${GITHUB_REF_NAME}"
 
     - name: Build production web artifact
       shell: sh

infra/docker/deploy-staging.sh

@@ -41,6 +41,13 @@ export COMPOSE_DOCKER_CLI_BUILD=0
 # Ensure isolated staging network exists
 docker network create xiaoxia-net-staging 2>/dev/null || true
 
+# 登录内网 Registry（拉取 API/Worker 镜像需要认证）
+REGISTRY_USER="${REGISTRY_USER:-admin}"
+REGISTRY_PASS="${REGISTRY_PASS:-}"
+if [ -n "$REGISTRY_PASS" ]; then
+  printf '%s' "$REGISTRY_PASS" | docker login "$REGISTRY" -u "$REGISTRY_USER" --password-stdin 2>/dev/null || true
+fi
+
 # Set default image names with registry prefix if not provided
 export API_IMAGE="${API_IMAGE:-${REGISTRY}/xiaoxia-saas-api:dev}"
 export WORKER_IMAGE="${WORKER_IMAGE:-${REGISTRY}/xiaoxia-saas-worker:dev}"

infra/docker/worker.Dockerfile

@@ -24,13 +24,14 @@ WORKDIR /app
 # ---- 依赖分层：基础依赖（变化少，缓存命中率高）----
 COPY requirements-base.txt /tmp/requirements-base.txt
 
-RUN pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com -r /tmp/requirements-base.txt \
+RUN python -m venv /opt/venv \
+    && /opt/venv/bin/pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com -r /tmp/requirements-base.txt \
     && rm /tmp/requirements-base.txt
 
 # ---- 依赖分层：业务依赖（变化频繁）----
 COPY requirements.txt /tmp/requirements.txt
 
-RUN pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com -r /tmp/requirements.txt \
+RUN /opt/venv/bin/pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com -r /tmp/requirements.txt \
     && rm /tmp/requirements.txt
 
 # 复制应用代码
@@ -42,6 +43,7 @@ COPY alembic.ini /app/alembic.ini
 COPY migrations/ /app/migrations/
 
 # 设置 Python 路径
+ENV PATH="/opt/venv/bin:$PATH"
 ENV PYTHONPATH=/app
 ENV PYTHONUNBUFFERED=1
 

scripts/build_release_images.sh

@@ -28,8 +28,43 @@ if docker ps --format '{{.Names}}' | grep -Eq '^(xiaoxia-(api|web|worker|postgre
   fi
 fi
 
-docker build --pull=false -f infra/docker/api.Dockerfile -t "$API_IMAGE" -t "$API_LATEST" .
-docker build --pull=false -f infra/docker/worker.Dockerfile -t "$WORKER_IMAGE" -t "$WORKER_LATEST" .
+# ---- BuildKit 缓存优化 ----
+CACHE_REGISTRY="${CACHE_REGISTRY:-git.xiaoxiajianji.com/xiaoxia/xiaoxia-saas}"
+CACHE_TAG="${CACHE_TAG:-release}"
+USE_CACHE=0
+
+# 检查 buildx 和 Gitea Registry 认证是否可用
+if docker buildx version >/dev/null 2>&1; then
+  # 尝试登录缓存 Registry（有 token 才启用缓存）
+  if [ -n "${REGISTRY_TOKEN:-}" ]; then
+    printf '%s' "${REGISTRY_TOKEN}" | docker login git.xiaoxiajianji.com -u xiaoxia --password-stdin 2>/dev/null && USE_CACHE=1
+  fi
+  # 确保使用 docker driver（共享 daemon 凭证）
+  docker buildx use default 2>/dev/null || true
+fi
+
+if [ "$USE_CACHE" -eq 1 ]; then
+  echo "Using buildx with distributed cache (${CACHE_REGISTRY})"
+  docker buildx build \
+    --cache-from "type=registry,ref=${CACHE_REGISTRY}/api-cache:${CACHE_TAG},ignore-error=true" \
+    --cache-to "type=registry,ref=${CACHE_REGISTRY}/api-cache:${CACHE_TAG},mode=max" \
+    -f infra/docker/api.Dockerfile \
+    -t "$API_IMAGE" -t "$API_LATEST" \
+    --load \
+    .
+  docker buildx build \
+    --cache-from "type=registry,ref=${CACHE_REGISTRY}/worker-cache:${CACHE_TAG},ignore-error=true" \
+    --cache-to "type=registry,ref=${CACHE_REGISTRY}/worker-cache:${CACHE_TAG},mode=max" \
+    -f infra/docker/worker.Dockerfile \
+    -t "$WORKER_IMAGE" -t "$WORKER_LATEST" \
+    --load \
+    .
+else
+  echo "Buildx cache not available, using plain docker build"
+  docker build --pull=false -f infra/docker/api.Dockerfile -t "$API_IMAGE" -t "$API_LATEST" .
+  docker build --pull=false -f infra/docker/worker.Dockerfile -t "$WORKER_IMAGE" -t "$WORKER_LATEST" .
+fi
+
 docker save "$API_IMAGE" "$API_LATEST" "$WORKER_IMAGE" "$WORKER_LATEST" -o "$TAR_PATH"
 
 printf '%s\n' "$TAR_PATH"