fix(cd): 修復 CD Pipeline 11 項安全/可靠性問題
🔴 Critical: C1 commit message injection: 所有 ${{ }} 值改走 env: 區塊隔離,不直接嵌入 shell C2 SSH StrictHostKeyChecking: 改用 known_hosts 驗證,支援 SSH_HOST_KEY secret 🟠 High: H1 rsync excludes 對齊: Rebuild 模式補齊 .gitea/ .claude/ docs/ *.md 等 7 條 H2 --force-recreate: Rebuild 模式加入強制重建,防止靜默更新失敗 H3 健康檢查強化: 加入 SSH 驗三容器 Running 狀態(scheduler/telegram-bot) H4 緊急回滾: 部署失敗時自動嘗試 docker restart 三容器回復服務 H5 ADR-011 守衛: Rebuild 前確認 momo-postgres 存活才繼續 🟡 Medium: M1 .claude/ 加入 rsync excludes(不同步 hook 腳本至 188) M2 *.md 加入 rsync excludes(根目錄 markdown 不需同步) M3 workflow_dispatch 新增 force_rebuild boolean 輸入 M4 cancel-in-progress 已知風險記入檔頭注意事項 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -5,6 +5,13 @@
|
||||
# 部署架構: Docker Compose on ollama@192.168.0.188 (Volume Mount)
|
||||
# 加速措施: Python 檔案走 rsync,僅 Dockerfile/requirements 變動才重建 image
|
||||
# 參考: AWOOOI cd.yaml pattern (ADR-008 — Docker Compose 非 K8s)
|
||||
#
|
||||
# 安全注意:
|
||||
# SSH_HOST_KEY secret: 請執行 ssh-keyscan 192.168.0.188 並將輸出存入 Gitea Secret
|
||||
# 若未設定,自動 ssh-keyscan(私有網段可接受的降級)
|
||||
#
|
||||
# 已知風險:
|
||||
# cancel-in-progress: rsync 非原子,若新 push 在傳輸中取消,188 可能半更新狀態
|
||||
|
||||
name: CD Pipeline
|
||||
|
||||
@@ -36,7 +43,11 @@ on:
|
||||
- '.gitea/workflows/**'
|
||||
# docs/、memory/、ADR、k8s/ 等不觸發
|
||||
workflow_dispatch:
|
||||
# 手動觸發永遠可用(用於補跑、緊急部署)
|
||||
inputs:
|
||||
force_rebuild:
|
||||
description: '強制重建 Docker Image(不論變更檔案)'
|
||||
type: boolean
|
||||
default: false
|
||||
|
||||
# 新 push 立即取消舊 job,只部署最新版本
|
||||
concurrency:
|
||||
@@ -59,12 +70,16 @@ jobs:
|
||||
echo "short_sha=${GITHUB_SHA::7}" >> $GITHUB_OUTPUT
|
||||
echo "message=$(git log -1 --pretty=%s | head -c 60)" >> $GITHUB_OUTPUT
|
||||
echo "start_time=$(date +%s)" >> $GITHUB_OUTPUT
|
||||
echo "actor=${{ github.actor }}" >> $GITHUB_OUTPUT
|
||||
|
||||
# 偵測是否需重建 Docker image(Dockerfile / requirements.txt / docker-compose.yml 變動)
|
||||
# 偵測是否需重建 Docker image(force_rebuild 優先,其次看變更檔案)
|
||||
- name: 偵測部署類型
|
||||
id: deploy_type
|
||||
run: |
|
||||
if [[ "${{ github.event.inputs.force_rebuild }}" == "true" ]]; then
|
||||
echo "type=rebuild" >> $GITHUB_OUTPUT
|
||||
echo "label=🔨 強制重建 Docker Image" >> $GITHUB_OUTPUT
|
||||
exit 0
|
||||
fi
|
||||
CHANGED=$(git diff --name-only HEAD~1 HEAD 2>/dev/null || echo "")
|
||||
if echo "$CHANGED" | grep -qE '^(Dockerfile|requirements\.txt|docker-compose\.yml)$'; then
|
||||
echo "type=rebuild" >> $GITHUB_OUTPUT
|
||||
@@ -74,32 +89,46 @@ jobs:
|
||||
echo "label=📁 同步 Python 檔案" >> $GITHUB_OUTPUT
|
||||
fi
|
||||
|
||||
# 設定 SSH 金鑰(用於連線到 188)
|
||||
# 設定 SSH 金鑰 + 主機驗證(C2 fix: 移除 StrictHostKeyChecking no)
|
||||
- name: 設定 SSH 金鑰
|
||||
env:
|
||||
DEPLOY_SSH_KEY: ${{ secrets.DEPLOY_SSH_KEY }}
|
||||
SSH_HOST_KEY: ${{ secrets.SSH_HOST_KEY }}
|
||||
run: |
|
||||
mkdir -p ~/.ssh
|
||||
echo "${{ secrets.DEPLOY_SSH_KEY }}" > ~/.ssh/id_deploy
|
||||
printf '%s\n' "$DEPLOY_SSH_KEY" > ~/.ssh/id_deploy
|
||||
chmod 600 ~/.ssh/id_deploy
|
||||
# 主機驗證:優先使用 SSH_HOST_KEY secret,否則動態掃描(私有網段降級)
|
||||
if [[ -n "$SSH_HOST_KEY" ]]; then
|
||||
echo "$SSH_HOST_KEY" >> ~/.ssh/known_hosts
|
||||
else
|
||||
ssh-keyscan -H 192.168.0.188 >> ~/.ssh/known_hosts 2>/dev/null
|
||||
fi
|
||||
chmod 644 ~/.ssh/known_hosts
|
||||
cat > ~/.ssh/config << 'EOF'
|
||||
Host 192.168.0.188
|
||||
HostName 192.168.0.188
|
||||
User ollama
|
||||
IdentityFile ~/.ssh/id_deploy
|
||||
StrictHostKeyChecking no
|
||||
ConnectTimeout 10
|
||||
EOF
|
||||
|
||||
# 通知部署開始(C1 fix: 所有 ${{ }} 值改走 env: 區塊,不直接嵌入 shell)
|
||||
- name: 通知部署開始
|
||||
env:
|
||||
COMMIT_MSG: ${{ steps.commit.outputs.message }}
|
||||
COMMIT_SHA: ${{ steps.commit.outputs.short_sha }}
|
||||
COMMIT_ACTOR: ${{ github.actor }}
|
||||
DEPLOY_LABEL: ${{ steps.deploy_type.outputs.label }}
|
||||
TG_TOKEN: ${{ secrets.TELEGRAM_BOT_TOKEN }}
|
||||
TG_CHAT: ${{ secrets.TELEGRAM_CHAT_ID }}
|
||||
run: |
|
||||
COMMIT_ESC=$(echo "${{ steps.commit.outputs.message }}" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
COMMIT_ESC=$(printf '%s' "$COMMIT_MSG" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
MSG=$(printf '🚀 <b>EwoooC 部署開始</b>\n├ 📝 <code>%s</code>\n├ 🔖 <code>%s</code>\n├ 👤 %s\n└ %s' \
|
||||
"${COMMIT_ESC}" \
|
||||
"${{ steps.commit.outputs.short_sha }}" \
|
||||
"${{ steps.commit.outputs.actor }}" \
|
||||
"${{ steps.deploy_type.outputs.label }}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${{ secrets.TELEGRAM_BOT_TOKEN }}/sendMessage" \
|
||||
"${COMMIT_ESC}" "${COMMIT_SHA}" "${COMMIT_ACTOR}" "${DEPLOY_LABEL}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${TG_TOKEN}/sendMessage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d "$(jq -n --arg c "${{ secrets.TELEGRAM_CHAT_ID }}" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
-d "$(jq -n --arg c "$TG_CHAT" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
|
||||
# ── 安裝部署工具 ────────────────────────────────────────────────────
|
||||
- name: 安裝 rsync / ssh
|
||||
@@ -111,9 +140,10 @@ jobs:
|
||||
if: steps.deploy_type.outputs.type == 'sync'
|
||||
run: |
|
||||
rsync -avz \
|
||||
-e "ssh -i ~/.ssh/id_deploy -o StrictHostKeyChecking=no" \
|
||||
-e "ssh -i ~/.ssh/id_deploy" \
|
||||
--exclude='.git/' \
|
||||
--exclude='.gitea/' \
|
||||
--exclude='.claude/' \
|
||||
--exclude='data/' \
|
||||
--exclude='logs/' \
|
||||
--exclude='backups/' \
|
||||
@@ -125,6 +155,7 @@ jobs:
|
||||
--exclude='.env' \
|
||||
--exclude='*.db' \
|
||||
--exclude='*.db-journal' \
|
||||
--exclude='*.md' \
|
||||
--exclude='docs/' \
|
||||
--exclude='memory/' \
|
||||
--exclude='k8s/' \
|
||||
@@ -135,7 +166,7 @@ jobs:
|
||||
- name: 重啟容器(Sync 模式)
|
||||
if: steps.deploy_type.outputs.type == 'sync'
|
||||
run: |
|
||||
ssh -i ~/.ssh/id_deploy -o StrictHostKeyChecking=no ollama@192.168.0.188 \
|
||||
ssh -i ~/.ssh/id_deploy ollama@192.168.0.188 \
|
||||
"docker restart momo-pro-system momo-scheduler momo-telegram-bot 2>&1 && \
|
||||
echo '✅ 三容器已重啟(app/scheduler/telegram-bot)'"
|
||||
|
||||
@@ -143,10 +174,16 @@ jobs:
|
||||
- name: 同步所有檔案並重建 Image
|
||||
if: steps.deploy_type.outputs.type == 'rebuild'
|
||||
run: |
|
||||
# 先同步全部檔案
|
||||
# H5: ADR-011 守衛 — momo-postgres 必須存活才允許 rebuild
|
||||
ssh -i ~/.ssh/id_deploy ollama@192.168.0.188 \
|
||||
"docker ps --format '{{.Names}}' | grep -q 'momo-postgres' || \
|
||||
(echo 'ABORT: momo-postgres not running' && exit 1)"
|
||||
# H1: 與 Sync 模式對齊的完整 excludes(含 .gitea/ .claude/ docs/ *.md)
|
||||
rsync -avz \
|
||||
-e "ssh -i ~/.ssh/id_deploy -o StrictHostKeyChecking=no" \
|
||||
-e "ssh -i ~/.ssh/id_deploy" \
|
||||
--exclude='.git/' \
|
||||
--exclude='.gitea/' \
|
||||
--exclude='.claude/' \
|
||||
--exclude='data/' \
|
||||
--exclude='logs/' \
|
||||
--exclude='backups/' \
|
||||
@@ -157,17 +194,24 @@ jobs:
|
||||
--exclude='*.pyc' \
|
||||
--exclude='.env' \
|
||||
--exclude='*.db' \
|
||||
--exclude='*.db-journal' \
|
||||
--exclude='*.md' \
|
||||
--exclude='docs/' \
|
||||
--exclude='memory/' \
|
||||
--exclude='k8s/' \
|
||||
--exclude='n8n-workflows/' \
|
||||
--exclude='aiops-core/' \
|
||||
./ ollama@192.168.0.188:/home/ollama/momo-pro/
|
||||
# 重建並重啟(先強制移除舊容器,避免 container name conflict)
|
||||
ssh -i ~/.ssh/id_deploy -o StrictHostKeyChecking=no ollama@192.168.0.188 \
|
||||
# H2: --force-recreate 確保容器強制重建(避免靜默更新失敗)
|
||||
ssh -i ~/.ssh/id_deploy ollama@192.168.0.188 \
|
||||
"cd /home/ollama/momo-pro && \
|
||||
docker stop momo-pro-system momo-scheduler momo-telegram-bot 2>/dev/null; \
|
||||
docker rm momo-pro-system momo-scheduler momo-telegram-bot 2>/dev/null; \
|
||||
docker compose build momo-app && \
|
||||
docker compose up -d --no-deps momo-app scheduler telegram-bot && \
|
||||
docker compose up -d --no-deps --force-recreate momo-app scheduler telegram-bot && \
|
||||
echo '✅ Image 重建完成(三容器)'"
|
||||
|
||||
# ── 健康檢查(最多重試 5 次,每次間隔 10s) ───────────────────────────
|
||||
# ── 健康檢查(H3: HTTP + 三容器狀態雙重驗證) ─────────────────────────
|
||||
- name: 健康檢查
|
||||
run: |
|
||||
echo "⏳ 等待服務啟動(15s)..."
|
||||
@@ -175,19 +219,32 @@ jobs:
|
||||
for i in $(seq 1 5); do
|
||||
HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" https://mo.wooo.work/health --max-time 10 || echo "000")
|
||||
if [ "$HTTP_CODE" = "200" ]; then
|
||||
echo "✅ 健康檢查通過(HTTP $HTTP_CODE)"
|
||||
exit 0
|
||||
echo "✅ HTTP 健康檢查通過(HTTP $HTTP_CODE)"
|
||||
break
|
||||
fi
|
||||
echo "⏳ 嘗試 $i/5,HTTP $HTTP_CODE,等待 10s..."
|
||||
[ "$i" -eq 5 ] && echo "❌ HTTP 健康檢查失敗" && exit 1
|
||||
sleep 10
|
||||
done
|
||||
echo "❌ 健康檢查失敗"
|
||||
exit 1
|
||||
# 驗證三應用容器均在 Running 狀態
|
||||
ssh -i ~/.ssh/id_deploy ollama@192.168.0.188 \
|
||||
'RUNNING=$(docker ps --format "{{.Names}}" | grep -cE "momo-(pro-system|scheduler|telegram-bot)" || true); \
|
||||
if [ "$RUNNING" -lt 3 ]; then \
|
||||
docker ps --format "{{.Names}}\t{{.Status}}" | grep momo-; \
|
||||
echo "❌ 容器未全部就緒(Running: $RUNNING/3)"; exit 1; \
|
||||
else \
|
||||
echo "✅ 三容器均正常運行($RUNNING/3)"; \
|
||||
fi'
|
||||
|
||||
# ── 觸發 Post-Deploy Code Review ─────────────────────────────────────
|
||||
- name: 觸發 AI Code Review
|
||||
if: success()
|
||||
continue-on-error: true
|
||||
env:
|
||||
WEBHOOK_TOKEN: ${{ secrets.INTERNAL_WEBHOOK_TOKEN }}
|
||||
COMMIT_SHA_FULL: ${{ github.sha }}
|
||||
BRANCH_NAME: ${{ github.ref_name }}
|
||||
DEPLOY_TYPE: ${{ steps.deploy_type.outputs.type }}
|
||||
run: |
|
||||
CHANGED=$(git diff --name-only HEAD~1 HEAD 2>/dev/null || echo "")
|
||||
FILES_JSON=$(echo "$CHANGED" | grep -E '\.(py|yaml|yml|json)$' | \
|
||||
@@ -195,34 +252,55 @@ jobs:
|
||||
curl -fS --max-time 10 \
|
||||
-X POST "https://mo.wooo.work/code-review/api/internal/trigger" \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "X-Internal-Token: ${{ secrets.INTERNAL_WEBHOOK_TOKEN }}" \
|
||||
-d "{\"commit_sha\":\"${{ github.sha }}\",\"changed_files\":${FILES_JSON},\"branch\":\"${{ github.ref_name }}\",\"deploy_type\":\"${{ steps.deploy_type.outputs.type }}\"}" \
|
||||
-H "X-Internal-Token: ${WEBHOOK_TOKEN}" \
|
||||
-d "$(jq -n \
|
||||
--arg sha "$COMMIT_SHA_FULL" \
|
||||
--argjson files "$FILES_JSON" \
|
||||
--arg branch "$BRANCH_NAME" \
|
||||
--arg type "$DEPLOY_TYPE" \
|
||||
'{commit_sha:$sha,changed_files:$files,branch:$branch,deploy_type:$type}')" \
|
||||
&& echo "✅ Code Review Pipeline 已觸發" \
|
||||
|| echo "⚠️ Code Review webhook 呼叫失敗(不影響部署結果)"
|
||||
|
||||
# ── 部署成功通知 ──────────────────────────────────────────────────────
|
||||
# ── 部署成功通知(C1 fix: env: 區塊隔離)────────────────────────────
|
||||
- name: 通知部署成功
|
||||
if: success()
|
||||
env:
|
||||
COMMIT_MSG: ${{ steps.commit.outputs.message }}
|
||||
COMMIT_SHA: ${{ steps.commit.outputs.short_sha }}
|
||||
START_TIME: ${{ steps.commit.outputs.start_time }}
|
||||
TG_TOKEN: ${{ secrets.TELEGRAM_BOT_TOKEN }}
|
||||
TG_CHAT: ${{ secrets.TELEGRAM_CHAT_ID }}
|
||||
run: |
|
||||
END_TIME=$(date +%s)
|
||||
DURATION=$((END_TIME - ${{ steps.commit.outputs.start_time }}))
|
||||
COMMIT_ESC=$(echo "${{ steps.commit.outputs.message }}" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
DURATION=$((END_TIME - START_TIME))
|
||||
COMMIT_ESC=$(printf '%s' "$COMMIT_MSG" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
MSG=$(printf '✅ <b>EwoooC 部署成功</b>\n├ 📝 <code>%s</code>\n├ 🔖 <code>%s</code>\n├ ⏱ 耗時 %ss\n└ 🌐 https://mo.wooo.work' \
|
||||
"${COMMIT_ESC}" \
|
||||
"${{ steps.commit.outputs.short_sha }}" \
|
||||
"${DURATION}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${{ secrets.TELEGRAM_BOT_TOKEN }}/sendMessage" \
|
||||
"${COMMIT_ESC}" "${COMMIT_SHA}" "${DURATION}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${TG_TOKEN}/sendMessage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d "$(jq -n --arg c "${{ secrets.TELEGRAM_CHAT_ID }}" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
-d "$(jq -n --arg c "$TG_CHAT" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
|
||||
# ── 部署失敗通知 ──────────────────────────────────────────────────────
|
||||
- name: 通知部署失敗
|
||||
# ── H4: 緊急回滾嘗試(部署失敗時嘗試重啟三容器回復服務)───────────────
|
||||
- name: 緊急回滾嘗試
|
||||
if: failure()
|
||||
run: |
|
||||
COMMIT_ESC=$(echo "${{ steps.commit.outputs.message }}" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
echo "⚠️ 部署失敗,嘗試重啟三容器回復服務..."
|
||||
ssh -i ~/.ssh/id_deploy ollama@192.168.0.188 \
|
||||
"docker restart momo-pro-system momo-scheduler momo-telegram-bot 2>&1 || true" || true
|
||||
|
||||
# ── 部署失敗通知(C1 fix: env: 區塊隔離)────────────────────────────
|
||||
- name: 通知部署失敗
|
||||
if: failure()
|
||||
env:
|
||||
COMMIT_MSG: ${{ steps.commit.outputs.message }}
|
||||
COMMIT_SHA: ${{ steps.commit.outputs.short_sha }}
|
||||
TG_TOKEN: ${{ secrets.TELEGRAM_BOT_TOKEN }}
|
||||
TG_CHAT: ${{ secrets.TELEGRAM_CHAT_ID }}
|
||||
run: |
|
||||
COMMIT_ESC=$(printf '%s' "$COMMIT_MSG" | sed 's/&/\&/g; s/</\</g; s/>/\>/g')
|
||||
MSG=$(printf '❌ <b>EwoooC 部署失敗</b>\n├ 📝 <code>%s</code>\n├ 🔖 <code>%s</code>\n└ 🔍 請查看 Gitea Actions 日誌' \
|
||||
"${COMMIT_ESC}" \
|
||||
"${{ steps.commit.outputs.short_sha }}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${{ secrets.TELEGRAM_BOT_TOKEN }}/sendMessage" \
|
||||
"${COMMIT_ESC}" "${COMMIT_SHA}")
|
||||
curl -fS -X POST "https://api.telegram.org/bot${TG_TOKEN}/sendMessage" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d "$(jq -n --arg c "${{ secrets.TELEGRAM_CHAT_ID }}" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
-d "$(jq -n --arg c "$TG_CHAT" --arg t "$MSG" '{chat_id:$c,text:$t,parse_mode:"HTML"}')"
|
||||
|
||||
Reference in New Issue
Block a user