13、Argo Rollouts请求分析和渐进式交付

请求分析和渐进式交付

Argo Rollouts中的分析（Analysis）是用于根据阶段性交付效果的测量结果来推动渐进式交付的机制
   ◼ 分析机制通过分析模板（AnalysisTemplate CRD）定义，而后在Rollout中调用它
   ◼ 运行某次特定的交付过程时，Argo Rollouts会将该Rollout调用的AnalysisTemplate实例化为AnalysisRun（CRD）

AnalysisTemplate资源与AnalysisRun CRD

示例说明
   ◼ 定义了一个名为service-name的参数
      ◆用于传递要测量的服务的名称
   ◼ 定义了success-rate指标
      ◆测试指定服务的请求成功率
   ◼ 每隔20s查询一次prometheus服务，共执行三次
   ◼ 成功条件为：第1个指标（下标为0）的结果值大于等于0.95
   ◼ 遇到3次错误即终止进一步的测试

AnalysisRun CRD
   ◼ 配置格式与AnalysisTemplaste大致相同，所不同的是，AnalysisRun用于调用并实例化分析模板

结合Prometheus指标进行analysis

一：让ingress-controller能够允许prometheus过来抓指标

修改ingress-nginx-controller

二：部署Prometheus

克隆仓库，进行部署
git clone https://github.com/iKubernetes/k8s-prom.git

[root@ubuntu2004 ~]#cd k8s-prom/
[root@ubuntu2004 k8s-prom]#ls
k8s-prometheus-adapter  namespace.yaml  podinfo     prometheus-ingress
kube-state-metrics      node_exporter   prometheus  README.md

创建名称空间
[root@ubuntu2004 k8s-prom]#cat namespace.yaml 
---
apiVersion: v1
kind: Namespace
metadata:
  name: prom

[root@ubuntu2004 k8s-prom]#kubectl apply -f namespace.yaml 

把prometheus目录下的资源进行部署（相关文件内容不在展示，请克隆自行查看）
[root@ubuntu2004 k8s-prom]#kubectl apply -f prometheus/ -n prom

查看prom名称空间下的pod
[root@ubuntu2004 k8s-prom]#kubectl get pods -n prom

三：让prometheus能够在集群外部进行访问，使用ingress把其发布出去

[root@ubuntu2004 k8s-prom]#kubectl apply -f prometheus-ingress -n prom

在桌面系统对域名进行解析，然后打开页面

外部请求都是正确的

再起一终端修改外部请求，如修改请求到错误的Url上

while true; do curl hello.magedu.com/magedu.com; sleep $[$RAMDOM%20]; done

四：拆掉之前的rollouts资源，建立基于prometheus监控，自动执行更新的rollouts资源

[root@ubuntu2004 ~]#cd learning-jenkins-cicd/09-argocd-and-rollout/rollout-demos/
[root@ubuntu2004 rollout-demos]#ls
01-basic-rollouts-demo.yaml                           04-rollouts-bluegreen-demo.yaml
02-rollouts-with-ingress-nginx-traffic-shifting.yaml  05-rollouts-bluegreen-with-analysis.yaml
03-rollouts-with-prometheus-analysis.yaml             06-argocd-application-and-rollouts.yaml

拆除
[root@ubuntu2004 rollout-demos]#kubectl delete -f 02-rollouts-with-ingress-nginx-traffic-shifting.yaml

取消所有的外部访问，不取消会产生大量的错误

部署新的示例
[root@ubuntu2004 rollout-demos]#cat 03-rollouts-with-prometheus-analysis.yaml 
# CopyRight: MageEdu <mage@magedu.com>
#
---
apiVersion: argoproj.io/v1alpha1
kind: AnalysisTemplate
metadata:
  name: success-rate
spec:
  args:
  - name: service-name
  - name: namespace
  metrics:
  - name: success-rate
    # NOTE: prometheus queries return results in the form of a vector.
    # So it is common to access the index 0 of the returned array to obtain the value
    successCondition: result[0] >= 0.95
    interval: 20s 
    count: 5
    failureLimit: 3
    provider:
      prometheus:
        address: http://prometheus.prom.svc.cluster.local:9090
        query: |
          sum(irate(nginx_ingress_controller_requests{service=~"{{args.service-name}}",namespace=~"{{args.namespace}}",status!~"[4-5].*"}[1m])) / 
          sum(irate(nginx_ingress_controller_requests{service=~"{{args.service-name}}",namespace=~"{{args.namespace}}"}[1m]))
---
apiVersion: argoproj.io/v1alpha1
kind: Rollout
metadata:
  name: rollouts-helloworld-with-traffic-shifting
spec:
  replicas: 10 
  strategy:
    canary:
      analysis:
        templates:
        - templateName: success-rate
        args:
        - name: service-name
          value: spring-boot-helloworld
        - name: namespace
          value: default
      canaryService: spring-boot-helloworld-canary
      stableService: spring-boot-helloworld
      trafficRouting:
        nginx:
          stableIngress: spring-boot-helloworld
      steps:
      - setCanaryScale:
          matchTrafficWeight: true
      - setWeight: 5
      - pause: {duration: 2m}
      - setWeight: 10
      - pause: {duration: 1m}
      - setWeight: 20
      - pause: {duration: 40}
      - setWeight: 40
      - pause: {duration: 20}
      - setWeight: 60
      - pause: {duration: 20}
      - setWeight: 80
      - pause: {duration: 20}
  revisionHistoryLimit: 5
  selector:
    matchLabels:
      app: spring-boot-helloworld
  template:
    metadata:
      labels:
        app: spring-boot-helloworld
    spec:
      containers:
      - name: spring-boot-helloworld
        image: ikubernetes/spring-boot-helloworld:v0.9.2
        ports:
        - name: http
          containerPort: 80
          protocol: TCP
        resources:
          requests:
            memory: 32Mi
            cpu: 50m
        livenessProbe:
          httpGet:
            path: '/'
            port: 80
            scheme: HTTP
          initialDelaySeconds: 3
        readinessProbe:
          httpGet:
            path: '/'
            port: 80
            scheme: HTTP
          initialDelaySeconds: 5
---
apiVersion: v1
kind: Service
metadata:
  name: spring-boot-helloworld
spec:
  ports:
  - port: 80
    targetPort: http
    protocol: TCP
    name: http
  selector:
    app: spring-boot-helloworld
---
apiVersion: v1
kind: Service
metadata:
  name: spring-boot-helloworld-canary
spec:
  ports:
  - port: 80
    targetPort: http
    protocol: TCP
    name: http
  selector:
    app: spring-boot-helloworld
---
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: spring-boot-helloworld
spec:
  ingressClassName: "nginx"
  rules:
    - host: hello.magedu.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: spring-boot-helloworld
                port:
                  number: 80
---

kubectl apply -f 03-rollouts-with-prometheus-analysis.yaml

部署完成之后，继续发起外部的访问请求

再打开一终端，进行非正常访问

进行版本更新

进行分析，查询5次，得平均值，如果满足>0.95得条件，则会自动更新，反之，会回滚