|
18544
|
13624
|
6
|
5
|
98cc8da660b8d4dba9887432490471d976c03f5f
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776011897
|
1776011891
|
1776011897
|
|
0
|
|
0
|
Edit
Delete
|
|
18554
|
13625
|
6
|
5
|
98cc8da660b8d4dba9887432490471d976c03f5f
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
15793
|
4
|
1776011907
|
1776011907
|
1776011898
|
1776011907
|
|
1
|
|
0
|
Edit
Delete
|
|
18802
|
13817
|
6
|
5
|
d2c68b13960de626f7a8d496bf1977d263eb7931
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776067830
|
1776066558
|
1776067830
|
|
1
|
|
0
|
Edit
Delete
|
|
18820
|
13823
|
6
|
5
|
d2c68b13960de626f7a8d496bf1977d263eb7931
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776067929
|
1776067831
|
1776067929
|
|
0
|
|
0
|
Edit
Delete
|
|
18831
|
13825
|
6
|
5
|
d2c68b13960de626f7a8d496bf1977d263eb7931
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
16042
|
4
|
1776067943
|
1776067943
|
1776067930
|
1776067943
|
|
1
|
|
0
|
Edit
Delete
|
|
19239
|
14176
|
6
|
5
|
204e3356f50776130b4976cf96f4deedfe36ab5f
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776170158
|
1776170082
|
1776170158
|
|
0
|
|
0
|
Edit
Delete
|
|
19284
|
14181
|
6
|
5
|
551c4d2e9b42cd14481ec48c3b2e2526cab4d58c
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776173356
|
1776170159
|
1776173356
|
|
0
|
|
0
|
Edit
Delete
|
|
19338
|
14196
|
6
|
5
|
551c4d2e9b42cd14481ec48c3b2e2526cab4d58c
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776173486
|
1776173358
|
1776173486
|
|
1
|
|
0
|
Edit
Delete
|
|
19361
|
14201
|
6
|
5
|
551c4d2e9b42cd14481ec48c3b2e2526cab4d58c
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776173494
|
1776173487
|
1776173494
|
|
0
|
|
0
|
Edit
Delete
|
|
19371
|
14202
|
6
|
5
|
551c4d2e9b42cd14481ec48c3b2e2526cab4d58c
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
16481
|
4
|
1776173504
|
1776173504
|
1776173496
|
1776173505
|
|
1
|
|
0
|
Edit
Delete
|
|
19567
|
14348
|
6
|
5
|
110abcc02b429bfac3ebe16a02a876c0ba2f4f62
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776215622
|
1776215542
|
1776215622
|
|
1
|
|
0
|
Edit
Delete
|
|
19585
|
14351
|
6
|
5
|
110abcc02b429bfac3ebe16a02a876c0ba2f4f62
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776215626
|
1776215623
|
1776215626
|
|
0
|
|
0
|
Edit
Delete
|
|
19595
|
14352
|
6
|
5
|
110abcc02b429bfac3ebe16a02a876c0ba2f4f62
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776215628
|
1776215627
|
1776215628
|
|
0
|
|
0
|
Edit
Delete
|
|
19605
|
14353
|
6
|
5
|
110abcc02b429bfac3ebe16a02a876c0ba2f4f62
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
16676
|
4
|
1776215639
|
1776215639
|
1776215630
|
1776215639
|
|
1
|
|
0
|
Edit
Delete
|
|
19724
|
14429
|
6
|
5
|
fd1878b707f31b05ee314173ac91491adb28bc30
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776237054
|
1776236986
|
1776237054
|
|
1
|
|
0
|
Edit
Delete
|
|
19739
|
14432
|
6
|
5
|
fd1878b707f31b05ee314173ac91491adb28bc30
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776237060
|
1776237055
|
1776237060
|
|
0
|
|
0
|
Edit
Delete
|
|
19749
|
14433
|
6
|
5
|
fd1878b707f31b05ee314173ac91491adb28bc30
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
16792
|
4
|
1776237070
|
1776237070
|
1776237061
|
1776237070
|
|
1
|
|
0
|
Edit
Delete
|
|
20318
|
14945
|
6
|
5
|
7b47df3186db279cfc071517a6c034aa213d926d
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776388769
|
1776388658
|
1776388769
|
|
1
|
|
0
|
Edit
Delete
|
|
20332
|
14947
|
6
|
5
|
7b47df3186db279cfc071517a6c034aa213d926d
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1776388773
|
1776388770
|
1776388773
|
|
0
|
|
0
|
Edit
Delete
|
|
20342
|
14948
|
6
|
5
|
7b47df3186db279cfc071517a6c034aa213d926d
|
0
|
部署到 Staging
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
17357
|
4
|
1776388783
|
1776388783
|
1776388774
|
1776388783
|
|
1
|
|
0
|
Edit
Delete
|
|
23245
|
17732
|
6
|
5
|
f8c99c0bccfc306a4dcc9afd03e1247a4ebd4a97
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1777219016
|
1777218745
|
1777219016
|
|
0
|
|
0
|
Edit
Delete
|
|
23290
|
17737
|
6
|
5
|
bd8162cf2c08d175a1cab0f69d42b648bb825c10
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1777219593
|
1777219017
|
1777219593
|
|
0
|
|
0
|
Edit
Delete
|
|
23336
|
17743
|
6
|
5
|
6f30ee7954e6142227f44fbcb36106e5ce6018eb
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1777219998
|
1777219595
|
1777219998
|
|
0
|
|
0
|
Edit
Delete
|
|
23381
|
17748
|
6
|
5
|
c04c1f61a7ca64acf31ed914678656d87ed6e35f
|
0
|
部署到 Staging
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-staging:
name: 部署到 Staging
runs-on: ubuntu-latest
if: >-
needs.gate.outputs.target_env == 'staging' || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.STAGING_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo "${{ secrets.STAGING_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
- name: 部署到 Staging
run: |
# 通过 ssh 环境变量传递，避免 heredoc 展开泄露 secret
ssh -o SendEnv=DEPLOY_PATH,API_IMAGE,FRONTEND_IMAGE,VERSION $USER@$HOST << 'EOF'
cd "$DEPLOY_PATH"
export API_IMAGE="$API_IMAGE"
export FRONTEND_IMAGE="$FRONTEND_IMAGE"
# 拉取最新镜像
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ $? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 滚动更新 API
docker compose -f docker-compose.prod.yml up -d --no-deps api
# 健康检查
RETRY=0
MAX_RETRY=12
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=$((RETRY + 1))
if [ $RETRY -ge $MAX_RETRY ]; then
echo "API 健康检查超时"
docker compose -f docker-compose.prod.yml logs --tail=50 api
exit 1
fi
echo " 等待 API 就绪... (${RETRY}/${MAX_RETRY})"
sleep 5
done
# 更新前端 + 重载 Nginx
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "$VERSION" > .deployed_version
echo "==> Staging 部署完成: $VERSION"
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Staging 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick
else
curl -sf http://localhost:3000/health || exit 1
echo "Staging 健康检查通过"
fi
EOF
env:
HOST: ${{ secrets.STAGING_HOST }}
USER: ${{ secrets.STAGING_USER }}
timeout-minutes: "15"
permissions:
contents: read
...
|
deploy-staging
|
["gate","build-and-push"]
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1777220527
|
1777220000
|
1777220527
|
|
0
|
|
0
|
Edit
Delete
|
|
9381
|
7509
|
6
|
5
|
ff3149170c6b0deb6d8151cb962592199b95bdd8
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774286322
|
1774286259
|
1774286322
|
|
0
|
|
0
|
Edit
Delete
|
|
9395
|
7511
|
6
|
5
|
ff3149170c6b0deb6d8151cb962592199b95bdd8
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774286327
|
1774286322
|
1774286327
|
|
0
|
|
0
|
Edit
Delete
|
|
9405
|
7512
|
6
|
5
|
ff3149170c6b0deb6d8151cb962592199b95bdd8
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8135
|
4
|
1774286340
|
1774286340
|
1774286327
|
1774286340
|
|
1
|
|
0
|
Edit
Delete
|
|
9544
|
7604
|
6
|
5
|
2ec5b7d8079ffd911c7b27a395d5aba3ceafe372
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774312505
|
1774312441
|
1774312505
|
|
0
|
|
0
|
Edit
Delete
|
|
9558
|
7606
|
6
|
5
|
2ec5b7d8079ffd911c7b27a395d5aba3ceafe372
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774312508
|
1774312506
|
1774312508
|
|
0
|
|
0
|
Edit
Delete
|
|
9568
|
7607
|
6
|
5
|
2ec5b7d8079ffd911c7b27a395d5aba3ceafe372
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8271
|
4
|
1774312521
|
1774312521
|
1774312509
|
1774312522
|
|
1
|
|
0
|
Edit
Delete
|
|
9620
|
7619
|
6
|
5
|
07680473f95a02e139e159147a93ef74e61f3db2
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774314829
|
1774314769
|
1774314829
|
|
0
|
|
0
|
Edit
Delete
|
|
9634
|
7621
|
6
|
5
|
07680473f95a02e139e159147a93ef74e61f3db2
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774314836
|
1774314830
|
1774314836
|
|
0
|
|
0
|
Edit
Delete
|
|
9644
|
7622
|
6
|
5
|
07680473f95a02e139e159147a93ef74e61f3db2
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8319
|
4
|
1774314849
|
1774314849
|
1774314837
|
1774314850
|
|
1
|
|
0
|
Edit
Delete
|
|
9691
|
7629
|
6
|
5
|
cfe1efeda7265f05374d3bd0036cf684a15f3cb9
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774315759
|
1774315758
|
1774315759
|
|
0
|
|
0
|
Edit
Delete
|
|
9701
|
7630
|
6
|
5
|
cfe1efeda7265f05374d3bd0036cf684a15f3cb9
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774315823
|
1774315760
|
1774315823
|
|
0
|
|
0
|
Edit
Delete
|
|
9715
|
7632
|
6
|
5
|
cfe1efeda7265f05374d3bd0036cf684a15f3cb9
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774315826
|
1774315824
|
1774315826
|
|
0
|
|
0
|
Edit
Delete
|
|
9725
|
7633
|
6
|
5
|
cfe1efeda7265f05374d3bd0036cf684a15f3cb9
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8362
|
4
|
1774315839
|
1774315839
|
1774315827
|
1774315840
|
|
1
|
|
0
|
Edit
Delete
|
|
9772
|
7640
|
6
|
5
|
8c39619c9cdb0d888d10942bf50533c8238021df
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774317028
|
1774316965
|
1774317028
|
|
0
|
|
0
|
Edit
Delete
|
|
9786
|
7642
|
6
|
5
|
8c39619c9cdb0d888d10942bf50533c8238021df
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774317033
|
1774317028
|
1774317033
|
|
0
|
|
0
|
Edit
Delete
|
|
9796
|
7643
|
6
|
5
|
8c39619c9cdb0d888d10942bf50533c8238021df
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8408
|
4
|
1774317048
|
1774317048
|
1774317033
|
1774317048
|
|
1
|
|
0
|
Edit
Delete
|
|
9847
|
7654
|
6
|
5
|
dbf34b08bbb60650d15b0c55262dbfe8d0a3a655
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774319102
|
1774319038
|
1774319102
|
|
0
|
|
0
|
Edit
Delete
|
|
9861
|
7656
|
6
|
5
|
dbf34b08bbb60650d15b0c55262dbfe8d0a3a655
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774319281
|
1774319102
|
1774319281
|
|
0
|
|
0
|
Edit
Delete
|
|
9872
|
7658
|
6
|
5
|
dbf34b08bbb60650d15b0c55262dbfe8d0a3a655
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8457
|
4
|
1774319297
|
1774319298
|
1774319281
|
1774319298
|
|
1
|
|
0
|
Edit
Delete
|
|
9924
|
7670
|
6
|
5
|
db7f39e63151b9c065646855287b8be73e13649b
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774321784
|
1774321622
|
1774321784
|
|
0
|
|
0
|
Edit
Delete
|
|
9938
|
7672
|
6
|
5
|
db7f39e63151b9c065646855287b8be73e13649b
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774321883
|
1774321785
|
1774321883
|
|
0
|
|
0
|
Edit
Delete
|
|
9949
|
7674
|
6
|
5
|
db7f39e63151b9c065646855287b8be73e13649b
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8507
|
4
|
1774321899
|
1774321900
|
1774321883
|
1774321900
|
|
1
|
|
0
|
Edit
Delete
|
|
10004
|
7689
|
6
|
5
|
81e883dfff9283af39b3dd2aa30e25ae2119e8f0
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774325277
|
1774325215
|
1774325277
|
|
0
|
|
0
|
Edit
Delete
|
|
10018
|
7691
|
6
|
5
|
81e883dfff9283af39b3dd2aa30e25ae2119e8f0
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774325282
|
1774325277
|
1774325282
|
|
0
|
|
0
|
Edit
Delete
|
|
10028
|
7692
|
6
|
5
|
81e883dfff9283af39b3dd2aa30e25ae2119e8f0
|
0
|
部署到 Production
|
0
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
0
|
3
|
0
|
1774325283
|
1774325282
|
1774325283
|
|
0
|
|
0
|
Edit
Delete
|
|
10038
|
7693
|
6
|
5
|
81e883dfff9283af39b3dd2aa30e25ae2119e8f0
|
0
|
部署到 Production
|
1
|
name: CI/CD Deploy
"on":
# test-pipeli name: CI/CD Deploy
"on":
# test-pipeline 通过后自动触发（仅 main 分支）
workflow_run:
workflows: ["Test Pipeline"]
types: [completed]
branches: [main]
# 版本标签触发完整部署
push:
tags: ['v*']
# 手动触发
workflow_dispatch:
inputs:
environment:
description: '部署环境'
required: true
default: 'staging'
type: choice
options:
- staging
- production
- aliyun
- rollback-production
- rollback-aliyun
skip_tests:
description: '跳过测试（紧急修复）'
required: false
default: false
type: boolean
version:
description: '部署版本号（留空使用自动版本）'
required: false
type: string
env:
IMAGE_PREFIX: ${{ github.repository_owner }}/juhi
NODE_VERSION: "20"
PNPM_VERSION: "8"
REGISTRY: ghcr.io
jobs:
deploy-production:
name: 部署到 Production
runs-on: ubuntu-latest
if: >-
startsWith(github.ref, 'refs/tags/v') || needs.gate.outputs.target_env == 'production'
steps:
- uses: actions/checkout@v4
- name: 配置 SSH
uses: webfactory/ssh-agent@v0.8.0
with:
ssh-private-key: ${{ secrets.PRODUCTION_SSH_KEY }}
- name: 配置 SSH Known Hosts
run: |
mkdir -p ~/.ssh
echo "${{ secrets.PRODUCTION_SSH_KNOWN_HOSTS }}" >> ~/.ssh/known_hosts
chmod 644 ~/.ssh/known_hosts
- name: 同步部署文件
run: |
rsync -avz --delete \
docker-compose.prod.yml \
scripts/ \
deploy/ \
$USER@$HOST:$DEPLOY_PATH/
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
- name: 数据库备份
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
BACKUP_DIR="/opt/juhi/backups"
mkdir -p "$BACKUP_DIR"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="$BACKUP_DIR/pre_deploy_${TIMESTAMP}.sql"
docker compose -f docker-compose.prod.yml exec -T postgres \
pg_dump -U "${DB_USER:-juhi}" -d "${DB_NAME:-juhi_revops}" -Fc > "$BACKUP_FILE"
if [ $? -eq 0 ]; then
echo "备份完成: $(du -h "$BACKUP_FILE" | cut -f1)"
else
echo "备份失败"
exit 1
fi
find "$BACKUP_DIR" -name "pre_deploy_*.sql" -mtime +30 -delete 2>/dev/null || true
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- name: 部署到 Production
run: |
ssh $USER@$HOST << EOF
cd $DEPLOY_PATH
export API_IMAGE="${API_IMAGE}"
export FRONTEND_IMAGE="${FRONTEND_IMAGE}"
docker compose -f docker-compose.prod.yml pull api frontend
# 数据库迁移
echo "==> 数据库迁移..."
docker compose -f docker-compose.prod.yml --profile migrate run --rm migrate
if [ \$? -ne 0 ]; then
echo "数据库迁移失败"
exit 1
fi
# 记录部署历史
CURRENT_API=\$(docker inspect --format='{{.Config.Image}}' juhi-api 2>/dev/null || echo "none")
CURRENT_FE=\$(docker inspect --format='{{.Config.Image}}' juhi-frontend 2>/dev/null || echo "none")
echo "\$(date -Iseconds)|\${CURRENT_API}|\${CURRENT_FE}" >> .deploy-history
tail -20 .deploy-history > .deploy-history.tmp && mv .deploy-history.tmp .deploy-history
# 蓝绿部署
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=2 api
RETRY=0
MAX_RETRY=15
until curl -sf http://localhost:3000/health > /dev/null 2>&1; do
RETRY=\$((RETRY + 1))
if [ \$RETRY -ge \$MAX_RETRY ]; then
echo "健康检查超时，回滚到单实例"
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
exit 1
fi
sleep 5
done
docker compose -f docker-compose.prod.yml up -d --no-deps --scale api=1 api
docker compose -f docker-compose.prod.yml up -d --no-deps frontend
docker compose -f docker-compose.prod.yml exec -T nginx nginx -s reload 2>/dev/null || true
docker image prune -f
echo "${VERSION}" > .deployed_version
echo "==> Production 部署完成: ${VERSION}"
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
DEPLOY_PATH: /opt/juhi
API_IMAGE: ${{ needs.build-and-push.outputs.api-image }}
FRONTEND_IMAGE: ${{ needs.build-and-push.outputs.frontend-image }}
VERSION: ${{ needs.build-and-push.outputs.version }}
- name: Production 部署验证
run: |
ssh $USER@$HOST << 'EOF'
cd /opt/juhi
if [ -f "./scripts/post-deploy-verify.sh" ]; then
chmod +x ./scripts/post-deploy-verify.sh
./scripts/post-deploy-verify.sh --quick || exit 1
else
curl -sf http://localhost:3000/health || exit 1
UNHEALTHY=$(docker compose -f docker-compose.prod.yml ps --format json | grep -c '"unhealthy"' || true)
if [ "$UNHEALTHY" -gt 0 ]; then
docker compose -f docker-compose.prod.yml ps
exit 1
fi
echo "Production 验证通过"
fi
EOF
env:
HOST: ${{ secrets.PRODUCTION_HOST }}
USER: ${{ secrets.PRODUCTION_USER }}
- if: startsWith(github.ref, 'refs/tags/v')
name: 创建 GitHub Release
uses: softprops/action-gh-release@v2
with:
generate_release_notes: "true"
make_latest: "true"
timeout-minutes: "20"
...
|
deploy-production
|
["gate","build-and-push","depl ["gate","build-and-push","deploy-staging"]...
|
["ubuntu-latest"]
|
8560
|
4
|
1774325296
|
1774325296
|
1774325284
|
1774325296
|
|
1
|
|
0
|
Edit
Delete
|