Jenkis pipeline构建项目实践-编写podTemplate实现和k8s对接

微信公众号搜索 DevOps和k8s全栈技术 ，即可关注公众号，也可扫描文章最后的二维码关注公众号，每天会分享技术文章供大家阅读参考哈~

在完成前文的jenkins server 在k8s环境部署之后，本文我们来测试在k8s集群环境中的jenkins pipeline构建项目和更新，具体环境要求如下：
1、jenkins pipeline插件安装成功
2、要更新的应用已提前部署
3、Jenkins slave中需要有kubectl、svn、mvn客户端且环境变量设置准确
4、Jenkis slave需要能和master的api-server进行正常通信（这里为了简便，前文构建jenkins server的时候直接对defalut这个service account做rabc授权。其他可选方案是使用~/.kube/config文件实现客户端的授权认证）
5、需要配置共享卷(slave编译完成之后把war包copy到共享卷位置)

一、创建构建任务

1、点击“jenkins ”——“新建 ”——“pipeline”，输入项目的名称，点击“OK”

2、下拉到“pipeline”配置，填写脚本

具体脚本内容如下：

podTemplate(name: 'jenkins-slave', cloud: 'kubernetes',
  namespace: 'default', label: 'jenkins-slave',
  serviceAccount: 'default', containers: [
  containerTemplate(
      name: 'jenkins-slave',
      image: 'harbor.59iedu.com/fjhb/jenkins-slave-toolkit:2018-08-10-v1',
      args: '${computer.jnlpmac} ${computer.name}',
      ttyEnabled: true,
      privileged: false,
      alwaysPullImage:true,
      )
  ],
  volumes: [
    persistentVolumeClaim(mountPath: '/tmp/', claimName: 'tomcat-jcsj-data')
  ]) {
  node('jenkins-slave') {
    stage('svn-checkout') {
      container('jnlp') {
          sh """
          svn checkout --username=yanglw --password=mypassword http://192.168.1.81/svn/fjhbjsb/k8s-pipeline-test --non-interactive 
          """
      }
    }

    stage('mvn-package') {
      container('jnlp') {
          sh """
          mkdir -p /var/jenkins_home && cd k8s-pipeline-test && mvn clean package && cp -rpf target/*.war /tmp/
          """
      }
    }

    stage('restart') {
      container('jnlp') {
          sh """
          pod_name=`kubectl  get pods -l name=jcsj-dev -o name | cut -d"/" -f2`
          kubectl delete pod \$pod_name
          """
      }
    }

    }
}

3、脚本释义
podTemplate节指定创建pod的模板和环境

Name：为pod的名称前缀
Cloud：为构建pod的云环境，需要和前面新建的云环境名称一样
Namespace：创建pod所在的namespace
Label: 创建pod对应的标签
serviceAccount：pod使用sa，这里使用default, 所以前文创建jenkins master deployment的时候多创建了一个default-role,这样自动创建出来的的jenkins-slave具有对应的api权限

containerTemplate节

指定创建容器的模板，当云环境里面配置pod模板后，容器模板以云环境的配置为准。

Volumes节

配置jenkins-slave pod挂载的卷，当云环境里面配置pod模板后，挂载卷以云环境配置为准

stage节为具体的pipeline步骤

这里第一步进行svn代码迁出；
第二步进行编译，并把包传到共享卷上面；
第三步对pod进行删除，因为应用采用deployment方式部署，所以可以实现重新创建pod，达到更新的效果。

二、构建准备

在开始构建之前，有必要介绍一下jcsj-dev的环境，这个环境是需要提前部署好的。部署文件如下

# cat deploy.yml 
---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: jcsj-dev
  namespace: default
spec:
  replicas: 1
  template:
    metadata:
      labels:
        name: jcsj-dev
    spec:
      containers:
      - name: jcsj-dev
        image: tomcat:latest
        imagePullPolicy: IfNotPresent
        ports:
        - name: web
          containerPort: 8080
        volumeMounts:
        - mountPath: /usr/local/tomcat/webapps
          name: tomcat-jcsj-data
      volumes:
      - name: tomcat-jcsj-data
        persistentVolumeClaim:
          claimName: tomcat-jcsj-data
---
apiVersion: v1
kind: Service
metadata:
  name: jcsj-dev
spec:
  type: NodePort
  ports:
    - port: 8080
      targetPort: 8080
      nodePort: 8453
  selector:
name: jcsj-dev

# cat pv.yaml 
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: tomcat-jcsj-data
spec:
  capacity:
    storage: 5Gi 
  accessModes:
  - ReadWriteMany 
  nfs: 
    path: /home/tomcat_jcsj
    server: 192.168.115.6
  persistentVolumeReclaimPolicy: Recycle 

---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: tomcat-jcsj-data
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 5Gi

三、更新项目代码

更新index.jsp并提交

四、开始构建

点击“立即构建”，构建过程中可以通过控制台输出看具体的日志信息

构建过程中观察jenkins master的日志输出：

# kubectl logs -f jenkins-master-588b89c75f-ztvgm 
INFO: Started provisioning Kubernetes Pod Template from kubernetes with 1 executors. Remaining excess workload: 0
Aug 13, 2018 10:14:22 AM hudson.slaves.NodeProvisioner$2 run
INFO: Kubernetes Pod Template provisioning successfully completed. We have now 2 computer(s)
Aug 13, 2018 10:14:22 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Created Pod: jenkins-slave-gjjbh in namespace default
Aug 13, 2018 10:14:22 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for Pod to be scheduled (0/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:28 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Container is waiting jenkins-slave-gjjbh [jnlp]: ContainerStateWaiting(message=null, reason=ContainerCreating, additionalProperties={})
Aug 13, 2018 10:14:28 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for Pod to be scheduled (1/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:34 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (1/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:35 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (2/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:36 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (3/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:37 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (4/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:38 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (5/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:39 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (6/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:40 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (7/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:41 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (8/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:42 AM hudson.TcpSlaveAgentListener$ConnectionHandler run
INFO: Accepted JNLP4-connect connection #1 from /172.30.66.4:47408
Aug 13, 2018 10:14:42 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (9/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:43 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (10/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:44 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (11/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:45 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (12/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:46 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (13/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:47 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (14/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:48 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (15/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:49 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (16/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:50 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (17/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:51 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (18/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:52 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (19/100): jenkins-slave-gjjbh
Aug 13, 2018 10:14:53 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesLauncher launch
INFO: Waiting for agent to connect (20/100): jenkins-slave-gjjbh
Aug 13, 2018 10:15:12 AM org.csanchez.jenkins.plugins.kubernetes.pipeline.ContainerExecDecorator$1 doLaunch
INFO: Created process inside pod: [jenkins-slave-gjjbh], container: [jnlp] with pid:[-1]
Aug 13, 2018 10:15:21 AM org.csanchez.jenkins.plugins.kubernetes.pipeline.ContainerExecDecorator$1 doLaunch
INFO: Created process inside pod: [jenkins-slave-gjjbh], container: [jnlp] with pid:[-1]
Aug 13, 2018 10:16:04 AM org.csanchez.jenkins.plugins.kubernetes.pipeline.ContainerExecDecorator$1 doLaunch
INFO: Created process inside pod: [jenkins-slave-gjjbh], container: [jnlp] with pid:[-1]
Aug 13, 2018 10:16:06 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesSlave _terminate
INFO: Terminating Kubernetes instance for agent jenkins-slave-gjjbh
Aug 13, 2018 10:16:06 AM jenkins.slaves.DefaultJnlpSlaveReceiver channelClosed
WARNING: Computer.threadPoolForRemoting [#8] for jenkins-slave-gjjbh terminated
java.nio.channels.ClosedChannelException
        at org.jenkinsci.remoting.protocol.impl.ChannelApplicationLayer.onReadClosed(ChannelApplicationLayer.java:208)
        at org.jenkinsci.remoting.protocol.ApplicationLayer.onRecvClosed(ApplicationLayer.java:222)
        at org.jenkinsci.remoting.protocol.ProtocolStack$Ptr.onRecvClosed(ProtocolStack.java:832)
        at org.jenkinsci.remoting.protocol.FilterLayer.onRecvClosed(FilterLayer.java:287)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.onRecvClosed(SSLEngineFilterLayer.java:181)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.switchToNoSecure(SSLEngineFilterLayer.java:283)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.processWrite(SSLEngineFilterLayer.java:503)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.processQueuedWrites(SSLEngineFilterLayer.java:248)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.doSend(SSLEngineFilterLayer.java:200)
        at org.jenkinsci.remoting.protocol.impl.SSLEngineFilterLayer.doCloseSend(SSLEngineFilterLayer.java:213)
        at org.jenkinsci.remoting.protocol.ProtocolStack$Ptr.doCloseSend(ProtocolStack.java:800)
        at org.jenkinsci.remoting.protocol.ApplicationLayer.doCloseWrite(ApplicationLayer.java:173)
        at org.jenkinsci.remoting.protocol.impl.ChannelApplicationLayer$ByteBufferCommandTransport.closeWrite(ChannelApplicationLayer.java:311)
        at hudson.remoting.Channel.close(Channel.java:1295)
        at hudson.remoting.Channel.close(Channel.java:1263)
        at hudson.slaves.SlaveComputer.closeChannel(SlaveComputer.java:708)
        at hudson.slaves.SlaveComputer.access$800(SlaveComputer.java:96)
        at hudson.slaves.SlaveComputer$3.run(SlaveComputer.java:626)
        at jenkins.util.ContextResettingExecutorService$1.run(ContextResettingExecutorService.java:28)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)

Aug 13, 2018 10:16:06 AM org.csanchez.jenkins.plugins.kubernetes.pipeline.PodTemplateStepExecution$PodTemplateCallback finished
INFO: Removing pod template and deleting pod jenkins-slave-7gnwx from cloud kubernetes
Aug 13, 2018 10:16:06 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesSlave _terminate
INFO: Terminated Kubernetes instance for agent default/jenkins-slave-gjjbh
Aug 13, 2018 10:16:06 AM org.csanchez.jenkins.plugins.kubernetes.KubernetesSlave _terminate
INFO: Disconnected computer jenkins-slave-gjjbh
Terminated Kubernetes instance for agent default/jenkins-slave-gjjbh
Aug 13, 2018 10:16:06 AM org.csanchez.jenkins.plugins.kubernetes.pipeline.PodTemplateStepExecution$PodTemplateCallback finished
WARNING: Failed to delete pod for agent default/jenkins-slave-7gnwx: not found
Aug 13, 2018 10:16:07 AM org.jenkinsci.plugins.workflow.job.WorkflowRun finish
INFO: k8s-pipeline-test #1 completed: SUCCESS

五、验证结果

六、后记

本文仅介绍的pipeline方式构建项目的过程，实际应用中还需要解决若干问题，例如：

1、svn checkout步骤需要明文配置用户名和密码，存在安全性问题
2、 每次执行构建都需要全量迁出代码，效率不高
3、要更新的应用需要提前部署好，不能适配首次编译部署场景
4、对共享卷存在依赖，解耦问题需要解决

七、续

研究了pipeline的语法结合shell脚本，基本可以解决上述问题，以下是pipeline代码

podTemplate(name: 'jenkins-slave', cloud: 'kubernetes',
  namespace: 'default', label: 'jenkins-slave') 
  {
  node('jenkins-slave') {
    stage('svn-checkout') {
      container('jnlp') {
         checkout([$class: 'SubversionSCM', 
         additionalCredentials: [], 
         excludedCommitMessages: '', 
         excludedRegions: '', 
         excludedRevprop: '', 
         excludedUsers: '', 
         filterChangelog: false, 
         ignoreDirPropChanges: false, 
         includedRegions: '', 
         locations: [[cancelProcessOnExternalsFail: true, 
         credentialsId: 'ef21f2d3-cb39-468a-aa19-9175d96a365f', 
         depthOption: 'infinity', 
         ignoreExternalsOption: true, 
         local: '.', 
         remote: 'http://192.168.1.81/svn/fjhbjsb/k8s-pipeline-test']], 
         quietOperation: true, workspaceUpdater: [$class: 'UpdateUpdater']])
      }
    }

    stage('mvn-package') {
      container('jnlp') {
          sh """
          mvn clean package 
          """
      }
    }

    stage('build-docker-image') {
      container('jnlp') {
        sh '''
        IMAGE_NAME="harbor.59iedu.com/fjhb/jcsj:v$BUILD_NUMBER"   
        CONATINER_NAME="jcsj-dev" 
        echo "192.168.1.227 harbor.59iedu.com" >> /etc/hosts 
        sed -i  "s/BUILD_NUMBER/$BUILD_NUMBER/g" tomcat.yaml
        sed -i  "s/BUILD_NUMBER=1/BUILD_NUMBER=$BUILD_NUMBER/g" rolling-update.sh 
        docker login harbor.59iedu.com -u admin -p Harbor12345 
        docker build -t $IMAGE_NAME . 
        docker push $IMAGE_NAME
        '''

      }
    }

    stage('apply update') {
      container('jnlp') {
          sh '''
          NAMESPACE=default
          DEPLOYMENT=tomcat-jcsj
          REGISTRY=harbor.59iedu.com/fjhb/jcsj:v$BUILD_NUMBER
          FILE=tomcat.yaml       
          STATUS=$(kubectl get deployment -l k8s-app=$DEPLOYMENT -o name -n $NAMESPACE)
          if [ ! $STATUS ];then
          kubectl create -f $FILE
          else
          kubectl set image deployment/$DEPLOYMENT $DEPLOYMENT=$REGISTRY --namespace=$NAMESPACE
          fi    
          '''
      }
    }
    }
}

其中，svn-checkout节代码通过“流水线语法”工具生成，同时需要给default sa用户相应的权限,否则无法实现apply update节配置

# kubectl create clusterrolebinding k8s-admin --clusterrole=cluster-admin  --user=system:serviceaccount:default:default

文章转载地址如下:

https://blog.51cto.com/ylw6006/2159861

往期精彩文章

kubernetes全栈技术+企业案例演示【带你快速掌握和使用k8s】

突破运维和开发瓶颈、Python、k8s、DevOps转型一网打尽！

python运维开发实战-基础篇

python运维和开发实战-高级篇

python运维和开发实战-安装和创建Django项目

Docker公司禁止被列入美国"实体名单"的国家、企业使用

kubernetes面试题汇总

DevOps视频和资料免费领取

kubernetes技术分享-可用于企业内部培训

谈谈我的IT发展之路

kubernetes系列文章第一篇-k8s基本介绍

kubernetes系列文章第二篇-kubectl

了解pod和pod的生命周期-这一篇文章就够了

Kubernetes中部署MySQL高可用集群

Prometheus+Grafana+Alertmanager搭建全方位的监控告警系统-超详细文档

k8s1.18多master节点高可用集群安装-超详细中文官方文档

k8s中蓝绿部署、金丝雀发布、滚动更新汇总

运维常见问题汇总-tomcat篇

关于linux内核参数的调优，你需要知道

kubernetes挂载ceph rbd和cephfs

报警神器Alertmanager发送报警到多个渠道

jenkins+kubernetes+harbor+gitlab构建企业级devops平台

kubernetes网络插件-flannel篇

kubernetes网络插件-calico篇

kubernetes认证、授权、准入控制

限制不同的用户操作k8s资源

面试真题&技术资料免费领取-覆盖面超全~

Prometheus监控MySQL

Prometheus监控Nginx

Prometheus监控Tomcat

linux面试题汇总

测试通过storageclass动态生成pv

通过编写k8s的资源清单yaml文件部署gitlab服务

helm安装和使用-通过helm部署k8s应用

k8s基于Ingress-nginx实现灰度发布

k8s的Pod安全策略

Prometheus Operator-上篇-安装和使用篇

Prometheus Operator-下篇

通过kubeconfig登陆k8s的dashboard ui界面

通过token令牌登陆k8s dashboard ui界面 

kubernetes集群的etcd数据库详细介绍

Linux网络流量监控工具

kubernetes搭建EFK日志管理系统

prometheus operator监控k8s集群之外的haproxy组件         

kubernetes ConfigMap存储卷                                            

技术交流群

学无止境，了解更多关于kubernetes/docker/devops/openstack/openshift/linux/IaaS/PaaS相关内容，想要获取更多资料和免费视频，可按如下方式进入技术交流群

                            扫码加群????

微信：luckylucky421302

微信公众号

                                     长按指纹关注公众号????

                                       

                                       点击在看少个 bug????