DevOps搭建(十六)-Jenkins+K8s部署详细步骤

1、整体部署架构图

2、编写脚本

vi pipeline.yml

apiVersion: apps/v1
kind: Deployment
metadata:
  namespace: test
  name: pipeline
  labels:
    app: pipeline
spec:
  replicas: 2
  selector:
    matchLabels:
      app: pipeline
  template:
    metadata:
      labels:
        app: pipeline
    spec:
      containers:
        - name: pipeline
          image: 192.168.88.125:80/repo/devops-pipeline-test:v4.0.0
          #镜像拉取策略，不管内容变不变都重新拉取
          imagePullPolicy: Always
          ports:
            - containerPort: 8080
---
apiVersion: v1
kind: Service
metadata:
  namespace: test
  #注意：这里要和Deployment的名称一致
  name: pipeline
  labels:
    app: pipeline
spec:
  selector:
    #注意：这里要和Deployment的app名称一致
    app: pipeline
  ports:
    #Service的端口
    - port: 8081
      #容器内的端口
      targetPort: 8080
  type: NodePort
---
#apiVersion值必须写成networking.k8s.io/v1
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  namespace: test
  name: pipeline
spec:
  #可视化界面配置的ingress名称
  ingressClassName: ingress
  rules:
    #映射的域名
    - host: k8s.pipeline.com
      http:
        paths:
          #访问什么样的路径
          - path: /
            #匹配方式，前缀匹配，即匹配上path配置的斜杠就可以
            pathType: Prefix
            backend:
              #自定用哪个service
              service:
                name: pipeline
                port:
                  #映射的端口
                  number: 8081

3、配置daemon.json

要想从Harbor仓库中拉取镜像，首先要往主机和从机的daemon.json文件中添加Harbor的地址信息

"insecure-registries":["192.168.88.125:80"]，别忘了80端口

vi /etc/docker/daemon.json

添加完之后，重启下docker

systemctl restart docker

4、配置Docker私服信息

要拉取镜像，需要Harbor登录进行拉取，到Kuboard控制台对应的命名空间找到密文菜单，点击创建Secret。

填写对应的信息即可

复制上图中的复制命令在k8s主机上运行

docker login 192.168.88.125:80/ -u admin -p Harbor12345

5、测试脚本

kubectl apply -f pipeline.yml

浏览器请求服务正常返回，脚本测试通过。

http://192.168.88.126:31603/test/test

6、Jenkins整合Kubernetes

6.1、将pipeline.yml推送到k8smaster

先在SpringBoot项目下创建pipeline.yml文件，内容为步骤2中的脚本内容。

接着生成流水线脚本，目标服务器选择k8smaster服务器，文件写pipeline.yml文件。

将流水线脚本粘贴到Jenkinsfile对应的步骤中

6.2、Jenkinsfile增加远程执行kubectl命令

在Jenkinsfile文件中，流水线步骤增加如下增加远程执行kubectl命令步骤。

原始脚本

kubectl apply -f /usr/local/k8s/pipeline.yml
#强制刷新deployment，防止pipeline.yml没发生变化时不重新部署
kubectl rollout restart deployment pipeline -n test

生成流线脚本：

sshPublisher(publishers: [sshPublisherDesc(configName: 'k8smaster', transfers: [sshTransfer(cleanRemote: false, excludes: '', execCommand: 'kubectl apply -f /usr/local/k8s/pipeline.yml', execTimeout: 120000, flatten: false, makeEmptyDirs: false, noDefaultExcludes: false, patternSeparator: '[, ]+', remoteDirectory: '', remoteDirectorySDF: false, removePrefix: '', sourceFiles: '')], usePromotionTimestamp: false, useWorkspaceInPromotion: false, verbose: false)])

粘贴到Jenkinsfile文件对应的步骤中

完整的Jenkinsfile文件内容如下：

//所有的脚本命令都放在pipeline中
pipeline {
    //指定任务在哪个集群节点中执行
    agent any

    //声明全局变量，方便后面使用
    environment {
        harborUser = 'admin'
        harborPassword = 'Harbor12345'
        harborAddress = '192.168.88.125:80'
        harborRepo = 'repo'
    }

    stages {
        stage('拉取代码') {
            steps {
                echo '开始拉取git仓库代码……'
                checkout scmGit(branches: [[name: '$tag']], extensions: [], userRemoteConfigs: [[url: 'http://192.168.88.125:8929/root/devops-test.git']])
                echo '开始拉取git仓库代码完毕。'
            }
        }
        stage('构建项目') {
            steps {
                echo '开始通过maven构建项目……'
                sh '/var/jenkins_home/maven/bin/mvn clean package -DskipTests'
                echo '通过maven构建项目完毕'
            }
        }
        stage('代码检查') {
            steps {
                echo 'TODO 通过SonarQube做代码质量检测'
            }
        }
        stage('制作镜像') {
            steps {
                echo '通过Docker制作自定义镜像……'
                sh '''mv ./target/*.jar ./docker/
                docker build -t ${JOB_NAME}:${tag} ./docker/'''
                echo '通过Docker制作自定义镜像完毕'
            }
        }
        stage('推送镜像') {
            steps {
                echo '将自定义对象推送到Harbor仓库……'
                sh '''docker login -u ${harborUser} -p ${harborPassword} ${harborAddress}
                docker tag ${JOB_NAME}:${tag} ${harborAddress}/${harborRepo}/${JOB_NAME}:${tag}
                docker push ${harborAddress}/${harborRepo}/${JOB_NAME}:${tag}
                docker image prune -f'''
                echo '将自定义对象推送到Harbor仓库完成'
            }
        }
        stage('将yml文件传到k8smaster') {
            steps {
                echo '将yml文件传到k8smaster……'
                sshPublisher(publishers: [sshPublisherDesc(configName: 'k8smaster', transfers: [sshTransfer(cleanRemote: false, excludes: '', execCommand: '', execTimeout: 120000, flatten: false, makeEmptyDirs: false, noDefaultExcludes: false, patternSeparator: '[, ]+', remoteDirectory: '', remoteDirectorySDF: false, removePrefix: '', sourceFiles: 'pipeline.yml')], usePromotionTimestamp: false, useWorkspaceInPromotion: false, verbose: false)])
                echo '将yml文件传到k8smaster完毕'
            }
        }
		stage('远程执行k8smaster的kubectl命令') {
            steps {
                echo '远程执行k8smaster的kubectl命令……'
                sshPublisher(publishers: [sshPublisherDesc(configName: 'k8smaster', transfers: [sshTransfer(cleanRemote: false, excludes: '', execCommand: 'kubectl apply -f /usr/local/k8s/pipeline.yml', execTimeout: 120000, flatten: false, makeEmptyDirs: false, noDefaultExcludes: false, patternSeparator: '[, ]+', remoteDirectory: '', remoteDirectorySDF: false, removePrefix: '', sourceFiles: '')], usePromotionTimestamp: false, useWorkspaceInPromotion: false, verbose: false)])
                echo '远程执行k8smaster的kubectl命令完毕'
            }
        }
    }
}