1
apiVersion: apps/v1
2
kind: Deployment
3
metadata:
4
  name: myapp
5
  labels:
6
    app: myapp
7
spec:
8
  replicas: 3
9
  strategy:
10
    type: Recreate  # 指定重建部署策略
11
  selector:
12
    matchLabels:
13
      app: myapp
14
  template:
15
    metadata:
16
      labels:
17
        app: myapp
18
    spec:
19
      containers:
20
      - name: myapp
21
        image: myapp:v2.0
22
        ports:
23
        - containerPort: 8080

1
apiVersion: apps/v1
2
kind: Deployment
3
metadata:
4
  name: myapp
5
  labels:
6
    app: myapp
7
spec:
8
  replicas: 10
9
  strategy:
10
    type: RollingUpdate
11
    rollingUpdate:
12
      maxSurge: 1        # 每次最多新增1个实例（可用百分比，如 25%）
13
      maxUnavailable: 0  # 不允许任何实例不可用，保证服务持续可用
14
  selector:
15
    matchLabels:
16
      app: myapp
17
  template:
18
    metadata:
19
      labels:
20
        app: myapp
21
        version: v2.0
22
    spec:
23
      containers:
24
      - name: myapp
25
        image: myapp:v2.0
26
        ports:
27
        - containerPort: 8080
28
        # 就绪探针：确保容器准备好接收流量
29
        readinessProbe:
30
          httpGet:
31
            path: /health
32
            port: 8080
33
          initialDelaySeconds: 5
34
          periodSeconds: 10
35
          timeoutSeconds: 3
36
          successThreshold: 2
37
          failureThreshold: 3
38
        # 存活探针：检测容器是否存活
39
        livenessProbe:
40
          httpGet:
41
            path: /health
42
            port: 8080
43
          initialDelaySeconds: 30
44
          periodSeconds: 10

1
# 蓝色环境Deployment
2
apiVersion: apps/v1
3
kind: Deployment
4
metadata:
5
  name: myapp-blue  # 蓝色版本
6
spec:
7
  replicas: 3
8
  selector:
9
    matchLabels:
10
      app: myapp
11
      version: blue
12
  template:
13
    metadata:
14
      labels:
15
        app: myapp
16
        version: blue  # 版本标签
17
    spec:
18
      containers:
19
      - name: myapp
20
        image: myapp:v1.0
21
        ports:
22
        - containerPort: 8080
23

24
---
25
# 绿色环境Deployment
26
apiVersion: apps/v1
27
kind: Deployment
28
metadata:
29
  name: myapp-green  # 绿色版本
30
spec:
31
  replicas: 3
32
  selector:
33
    matchLabels:
34
      app: myapp
35
      version: green
36
  template:
37
    metadata:
38
      labels:
39
        app: myapp
40
        version: green  # 版本标签
41
    spec:
42
      containers:
43
      - name: myapp
44
        image: myapp:v2.0
45
        ports:
46
        - containerPort: 8080
47

48
---
49
# Service负载均衡器（通过修改selector切换流量）
50
apiVersion: v1
51
kind: Service
52
metadata:
53
  name: myapp-service
54
spec:
55
  selector:
56
    app: myapp
57
    version: blue  # 切换到green即可完成蓝绿切换
58
  ports:
59
  - protocol: TCP
60
    port: 80
61
    targetPort: 8080
62
  type: LoadBalancer

1
# 红色集群启动配置
2
Resources:
3
  RedLaunchTemplate:
4
    Type: AWS::EC2::LaunchTemplate
5
    Properties:
6
      LaunchTemplateName: myapp-red-v1
7
      LaunchTemplateData:
8
        ImageId: ami-red-v1
9
        InstanceType: t3.medium
10
        UserData:
11
          Fn::Base64: !Sub |
12
            #!/bin/bash
13
            docker run -d myapp:v1.0
14

15
  # 黑色集群启动配置
16
  BlackLaunchTemplate:
17
    Type: AWS::EC2::LaunchTemplate
18
    Properties:
19
      LaunchTemplateName: myapp-black-v2
20
      LaunchTemplateData:
21
        ImageId: ami-black-v2
22
        InstanceType: t3.medium
23
        UserData:
24
          Fn::Base64: !Sub |
25
            #!/bin/bash
26
            docker run -d myapp:v2.0
27

28
  # 负载均衡器（切换目标组即可完成红黑切换）
29
  ApplicationLoadBalancer:
30
    Type: AWS::ElasticLoadBalancingV2::LoadBalancer
31
    Properties:
32
      Name: myapp-alb
33
      Subnets:
34
        - subnet-abc123
35
        - subnet-def456

1
apiVersion: networking.istio.io/v1beta1
2
kind: VirtualService
3
metadata:
4
  name: myapp-canary
5
spec:
6
  hosts:
7
  - myapp.example.com
8
  http:
9
  - match:
10
    - headers:
11
        user-type:
12
          exact: internal  # 内部员工100%使用金丝雀版本
13
    route:
14
    - destination:
15
        host: myapp
16
        subset: v2-canary
17
      weight: 100
18
  - route:  # 普通用户5%流量到金丝雀版本
19
    - destination:
20
        host: myapp
21
        subset: v1-stable
22
      weight: 95  # 95%流量到稳定版本
23
    - destination:
24
        host: myapp
25
        subset: v2-canary
26
      weight: 5   # 5%流量到金丝雀版本
27

28
---
29
# 定义版本子集
30
apiVersion: networking.istio.io/v1beta1
31
kind: DestinationRule
32
metadata:
33
  name: myapp-versions
34
spec:
35
  host: myapp
36
  subsets:
37
  - name: v1-stable
38
    labels:
39
      version: v1.0
40
  - name: v2-canary
41
    labels:
42
      version: v2.0

1
apiVersion: networking.istio.io/v1beta1
2
kind: VirtualService
3
metadata:
4
  name: myapp-shadow
5
spec:
6
  hosts:
7
  - myapp.example.com
8
  http:
9
  - route:
10
    - destination:
11
        host: myapp
12
        subset: v1-production
13
      weight: 100  # 100%流量到生产版本，用户收到此响应
14
    mirror:  # 流量镜像配置
15
      host: myapp
16
      subset: v2-shadow
17
    mirrorPercentage:
18
      value: 100.0  # 镜像100%的流量到影子版本
19

20
---
21
apiVersion: networking.istio.io/v1beta1
22
kind: DestinationRule
23
metadata:
24
  name: myapp-shadow-rule
25
spec:
26
  host: myapp
27
  subsets:
28
  - name: v1-production
29
    labels:
30
      version: v1.0
31
  - name: v2-shadow
32
    labels:
33
      version: v2.0-shadow

1
apiVersion: apps/v1
2
kind: Deployment
3
metadata:
4
  name: myapp-shadow
5
spec:
6
  replicas: 3  # 影子版本实例数
7
  selector:
8
    matchLabels:
9
      app: myapp
10
      version: v2.0-shadow
11
  template:
12
    metadata:
13
      labels:
14
        app: myapp
15
        version: v2.0-shadow
16
    spec:
17
      containers:
18
      - name: myapp
19
        image: myapp:v2.0
20
        ports:
21
        - containerPort: 8080
22
        env:
23
        - name: MODE
24
          value: "shadow"  # 标记为影子模式
25
        - name: LOG_LEVEL
26
          value: "debug"   # 详细日志用于对比分析

1
from unleash import UnleashClient
2
import logging
3

4
# 初始化功能开关客户端
5
unleash_client = UnleashClient(
6
    url="https://unleash.example.com/api",
7
    app_name="myapp",
8
    custom_headers={'Authorization': ':spoiler[your-api-token]'}
9
)
10

11
def process_payment(user_id, amount):
12
    """支付处理函数"""
13

14
    # 构建用户上下文
15
    context = {
16
        "userId": user_id,
17
        "properties": {
18
            "region": get_user_region(user_id),
19
            "user_type": get_user_type(user_id)
20
        }
21
    }
22

23
    # 检查新支付引擎功能开关
24
    if unleash_client.is_enabled(
25
        "new-payment-engine",  # 功能开关名称
26
        context
27
    ):
28
        # 使用新支付引擎
29
        logging.info(f"用户 {user_id} 使用新支付引擎")
30
        return new_payment_processor.process(amount)
31
    else:
32
        # 使用旧支付引擎
33
        logging.info(f"用户 {user_id} 使用旧支付引擎")
34
        return legacy_payment_processor.process(amount)
35

36
def get_user_region(user_id):
37
    """获取用户地域"""
38
    # 实现逻辑...
39
    pass
40

41
def get_user_type(user_id):
42
    """获取用户类型（普通/VIP等）"""
43
    # 实现逻辑...
44
    pass

1
# Unleash功能开关配置示例
2
feature:
3
  name: "new-payment-engine"
4
  description: "新一代支付引擎"
5
  enabled: true
6
  strategies:
7
    # 策略1：先开放给内部员工
8
    - name: "userWithId"
9
      parameters:
10
        userIds: "employee-001,employee-002,employee-003"
11

12
    # 策略2：按地域逐步开放
13
    - name: "flexibleRollout"
14
      parameters:
15
        rollout: "25"  # 25%的用户
16
        stickiness: "userId"
17
        groupId: "payment-engine"
18
      constraints:
19
        - contextName: "region"
20
          operator: "IN"
21
          values: ["beijing", "shanghai"]
22

23
    # 策略3：只对VIP用户开放
24
    - name: "default"
25
      constraints:
26
        - contextName: "user_type"
27
          operator: "IN"
28
          values: ["vip", "premium"]

1
# A/B测试Nginx配置
2
upstream backend_a {
3
    server app-v1:8080;
4
}
5

6
upstream backend_b {
7
    server app-v2:8080;
8
}
9

10
# 基于Cookie的A/B测试分流
11
split_clients "${remote_addr}${http_user_agent}" $ab_variant {
12
    50%     "a";  # 50%用户到A组
13
    *       "b";  # 50%用户到B组
14
}
15

16
server {
17
    listen 80;
18
    server_name myapp.example.com;
19

20
    location / {
21
        # 设置A/B测试Cookie（有效期30天）
22
        add_header Set-Cookie "ab_variant=$ab_variant; Max-Age=2592000; Path=/";
23

24
        # 根据分组转发到不同的后端
25
        if ($ab_variant = "a") {
26
            proxy_pass http://backend_a;
27
        }
28
        if ($ab_variant = "b") {
29
            proxy_pass http://backend_b;
30
        }
31

32
        # 添加追踪头，方便后端记录
33
        proxy_set_header X-AB-Variant $ab_variant;
34
        proxy_set_header Host $host;
35
        proxy_set_header X-Real-IP $remote_addr;
36
    }
37
}

1
# Grafana监控面板配置示例
2
apiVersion: v1
3
kind: ConfigMap
4
metadata:
5
  name: grafana-dashboard-deployment
6
data:
7
  deployment-monitoring.json: |
8
    {
9
      "panels": [
10
        {
11
          "title": "请求成功率",
12
          "targets": [{
13
            "expr": "sum(rate(http_requests_total{status!~\"5..\"}[5m])) / sum(rate(http_requests_total[5m]))"
14
          }],
15
          "alert": {
16
            "conditions": [
17
              {
18
                "evaluator": {"params": [0.995]},
19
                "operator": {"type": "lt"},
20
                "query": {"params": ["A", "5m", "now"]}
21
              }
22
            ]
23
          }
24
        },
25
        {
26
          "title": "P99响应时间",
27
          "targets": [{
28
            "expr": "histogram_quantile(0.99, rate(http_request_duration_seconds_bucket[5m]))"
29
          }],
30
          "alert": {
31
            "conditions": [
32
              {"evaluator": {"params": [0.5]}, "operator": {"type": "gt"}}
33
            ]
34
          }
35
        },
36
        {
37
          "title": "错误率",
38
          "targets": [{
39
            "expr": "sum(rate(http_requests_total{status=~\"5..\"}[5m])) / sum(rate(http_requests_total[5m]))"
40
          }],
41
          "alert": {
42
            "conditions": [
43
              {"evaluator": {"params": [0.01]}, "operator": {"type": "gt"}}
44
            ]
45
          }
46
        },
47
        {
48
          "title": "CPU使用率",
49
          "targets": [{
50
            "expr": "avg(rate(container_cpu_usage_seconds_total[5m])) by (pod)"
51
          }]
52
        },
53
        {
54
          "title": "内存使用率",
55
          "targets": [{
56
            "expr": "avg(container_memory_working_set_bytes / container_spec_memory_limit_bytes) by (pod)"
57
          }]
58
        }
59
      ]
60
    }

1
# 发布后复盘报告
2

3
## 发布信息
4
- 发布时间：2024-11-29 02:00:00
5
- 发布版本：v2.5.0
6
- 发布策略：金丝雀发布
7
- 负责人员：张三、李四
8

9
## 发布过程时间线
10
| 时间 | 事件 | 负责人 | 备注 |
11
|------|------|--------|------|
12
| 02:00 | 开始发布 | 张三 | |
13
| 02:15 | 5%流量切换 | 张三 | |
14
| 02:45 | 发现错误率上升 | 监控系统 | 从0.1%上升到0.3% |
15
| 02:50 | 定位问题原因 | 李四 | 配置文件错误 |
16
| 03:00 | 修复配置 | 李四 | |
17
| 03:15 | 错误率恢复正常 | 监控系统 | |
18
| 03:30 | 继续推进到25% | 张三 | |
19
| 05:00 | 全量上线完成 | 张三 | |
20

21
## 遇到的问题
22
1. **配置文件错误**
23
   - 现象：错误率从0.1%上升到0.3%
24
   - 原因：数据库连接池配置错误
25
   - 解决：修正配置文件
26
   - 影响：影响了约500个请求
27

28
## 改进建议
29
1. 加强配置文件的自动化测试
30
2. 提前在预生产环境进行更充分的测试
31
3. 优化告警规则，缩短问题发现时间
32

33
## 总结
34
本次发布整体顺利，虽然遇到配置问题，但通过监控及时发现并快速解决。
35
建议后续加强配置管理和自动化测试。

1
name: 自动化部署流水线
2

3
on:
4
  push:
5
    branches: [ main ]
6

7
jobs:
8
  build-and-deploy:
9
    runs-on: ubuntu-latest
10

11
    steps:
12
    # 1. 代码检出
13
    - name: Checkout代码
14
      uses: actions/checkout@v3
15

16
    # 2. 运行自动化测试
17
    - name: 运行单元测试
18
      run: npm test
19

20
    - name: 运行集成测试
21
      run: npm run test:integration
22

23
    # 3. 构建Docker镜像
24
    - name: 构建镜像
25
      run: |
26
        docker build -t myapp:${{ github.sha }} .
27
        docker push myapp:${{ github.sha }}
28

29
    # 4. 部署到Kubernetes（金丝雀发布）
30
    - name: 部署金丝雀版本
31
      run: |
32
        kubectl set image deployment/myapp-canary \
33
          myapp=myapp:${{ github.sha }}
34

35
    # 5. 监控健康状态
36
    - name: 健康检查
37
      run: |
38
        ./scripts/health-check.sh
39

40
    # 6. 如果健康检查失败，自动回滚
41
    - name: 自动回滚
42
      if: failure()
43
      run: |
44
        kubectl rollout undo deployment/myapp-canary
45
        ./scripts/alert-team.sh "部署失败，已自动回滚"
46

47
    # 7. 逐步推进
48
    - name: 推进到50%流量
49
      if: success()
50
      run: ./scripts/progressive-rollout.sh 50

1
apiVersion: v1
2
kind: ConfigMap
3
metadata:
4
  name: prometheus-alerts
5
data:
6
  alerts.yml: |
7
    groups:
8
    - name: deployment_alerts
9
      interval: 30s
10
      rules:
11
      # 1. 错误率告警
12
      - alert: HighErrorRate
13
        expr: |
14
          sum(rate(http_requests_total{status=~"5.."}[5m]))
15
          /
16
          sum(rate(http_requests_total[5m])) > 0.01
17
        for: 2m
18
        labels:
19
          severity: critical
20
        annotations:
21
          summary: "错误率超过1%"
22
          description: "当前错误率: {{ $value | humanizePercentage }}"
23

24
      # 2. 延迟告警
25
      - alert: HighLatency
26
        expr: |
27
          histogram_quantile(0.99,
28
            rate(http_request_duration_seconds_bucket[5m])
29
          ) > 0.5
30
        for: 3m
31
        labels:
32
          severity: warning
33
        annotations:
34
          summary: "P99延迟超过500ms"
35
          description: "当前P99延迟: {{ $value }}s"
36

37
      # 3. 资源使用率告警
38
      - alert: HighCPUUsage
39
        expr: |
40
          avg(rate(container_cpu_usage_seconds_total[5m]))
41
          by (pod) > 0.8
42
        for: 5m
43
        labels:
44
          severity: warning
45
        annotations:
46
          summary: "CPU使用率超过80%"
47
          description: "Pod {{ $labels.pod }} CPU使用率: {{ $value | humanizePercentage }}"
48

49
      # 4. 版本对比告警（金丝雀发布专用）
50
      - alert: CanaryVersionUnhealthy
51
        expr: |
52
          (
53
            sum(rate(http_requests_total{version="v2.0",status=~"5.."}[5m]))
54
            /
55
            sum(rate(http_requests_total{version="v2.0"}[5m]))
56
          )
57
          >
58
          (
59
            sum(rate(http_requests_total{version="v1.0",status=~"5.."}[5m]))
60
            /
61
            sum(rate(http_requests_total{version="v1.0"}[5m]))
62
          ) * 1.5
63
        for: 3m
64
        labels:
65
          severity: critical
66
        annotations:
67
          summary: "金丝雀版本错误率异常"
68
          description: "新版本错误率比旧版本高50%以上"

1
-- 数据库变更脚本示例
2
-- 版本：v2.5.0
3
-- 作者：张三
4
-- 日期：2024-11-29
5
-- 说明：为用户表添加会员等级字段
6

7
-- 步骤1：添加新字段（允许NULL，保持向后兼容）
8
ALTER TABLE users ADD COLUMN membership_level VARCHAR(20) NULL;
9

10
-- 步骤2：为现有数据填充默认值
11
UPDATE users SET membership_level = 'basic' WHERE membership_level IS NULL;
12

13
-- 步骤3：添加索引（可选，根据查询需求决定）
14
CREATE INDEX idx_membership_level ON users(membership_level);
15

16
-- 回滚脚本（保存为rollback-v1.sql）
17
-- ALTER TABLE users DROP COLUMN membership_level;
18
-- DROP INDEX idx_membership_level ON users;