kubersphere 3.1 按官方文档构建的流水线，无法正常工作。

Mike

按官方文档

https://kubesphere.com.cn/docs/devops-user-guide/how-to-use/create-a-pipeline-using-graphical-editing-panel/

构建了一个流水线，无法正常启动

报错如下：

对应的 maven 容器已经拉起来了

请大神些解决下

zhu733756

@Rick rick大佬帮忙看看

johnniang

给出一个复现步骤呢。

Mike

johnniang

pipeline {
agent {
node {
label ‘base’
}

}
stages {
stage(‘stage-w4jl7’) {
agent none
steps {
sh ‘echo “this is test”’
}
}

}

现在将流水线改成最简单的了还是报错

从报错看是 jenkins agent 连不上

但对应的 pod 是启动了的

应该还是节点没启动起来是什么原因呢？

看容器里的日志 jenkins agents 是连上了的

但jenkins 页面上又是报错的

而且容器里连的地址好像和 jenkins 页面上的并不一样？

johnniang

Mike 好的，我尝试复现一下。

Mike

johnniang 感谢，上面又更新了一些报错可以参考下

Mike

该问题已解决，原因是内置 pod template 里的镜像有问题

以 go 环境为例

将原来的 kubesphere/builder-go:v3.1.0 换成 kubesphere/builder-go:v2.1.0 就一切工作正常了

Rick

该问题已解决，原因是内置 pod template 里的镜像有问题。以 go 环境为例
我刚刚测试，builder-go:v3.1.0这个镜像是没问题的，请看下面的测试过程：

下面是测试流水线：

pipeline {
    agent {
     label 'go'
    
    }
    stages {
        stage('ss') {
            steps {
                container('go') {
                    sh 'go version'
                }
            }
        }
    
    }
}

下面是测试的输出日志：

Started by user admin
Running in Durability level: MAX_SURVIVABILITY
[Pipeline] Start of Pipeline
[Pipeline] node
Agent go-j8gf2 is provisioned from template go
---
apiVersion: "v1"
kind: "Pod"
metadata:
  annotations: {}
  labels:
    jenkins: "slave"
    jenkins/label-digest: "e84a3b2cb16af20654b162562a7360c9fc00fd54"
    jenkins/label: "go"
  name: "go-j8gf2"
spec:
  affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - preference:
          matchExpressions:
          - key: "node-role.kubernetes.io/worker"
            operator: "In"
            values:
            - "ci"
        weight: 1
  containers:
  - command:
    - "cat"
    image: "kubesphere/builder-go:v3.1.0"
    imagePullPolicy: "IfNotPresent"
    name: "go"
    resources:
      limits:
        ephemeral-storage: "10Gi"
        memory: "8192Mi"
        cpu: "4000m"
      requests:
        ephemeral-storage: "1Gi"
        memory: "100Mi"
        cpu: "100m"
    tty: true
    volumeMounts:
    - mountPath: "/home/jenkins/agent"
      name: "workspace-volume"
      readOnly: false
    - mountPath: "/root/.sonar/cache"
      name: "volume-2"
      readOnly: false
    - mountPath: "/var/run/docker.sock"
      name: "volume-0"
      readOnly: false
    - mountPath: "/home/jenkins/go/pkg"
      name: "volume-1"
      readOnly: false
    workingDir: "/home/jenkins/agent"
  - args:
    - "********"
    - "go-j8gf2"
    command:
    - "jenkins-slave"
    env:
    - name: "JENKINS_SECRET"
      value: "********"
    - name: "JENKINS_TUNNEL"
      value: "ks-jenkins-agent.kubesphere-devops-system:50000"
    - name: "JENKINS_AGENT_NAME"
      value: "go-j8gf2"
    - name: "JENKINS_NAME"
      value: "go-j8gf2"
    - name: "JENKINS_AGENT_WORKDIR"
      value: "/home/jenkins/agent"
    - name: "JENKINS_URL"
      value: "http://ks-jenkins.kubesphere-devops-system:80/"
    image: "registry.cn-beijing.aliyuncs.com/kubesphereio/jnlp-slave:3.27-1"
    imagePullPolicy: "IfNotPresent"
    name: "jnlp"
    resources:
      limits:
        memory: "1536Mi"
      requests:
        memory: "400Mi"
        cpu: "50m"
    tty: false
    volumeMounts:
    - mountPath: "/root/.sonar/cache"
      name: "volume-2"
      readOnly: false
    - mountPath: "/var/run/docker.sock"
      name: "volume-0"
      readOnly: false
    - mountPath: "/home/jenkins/go/pkg"
      name: "volume-1"
      readOnly: false
    - mountPath: "/home/jenkins/agent"
      name: "workspace-volume"
      readOnly: false
    workingDir: "/home/jenkins/agent"
  hostNetwork: false
  nodeSelector: {}
  restartPolicy: "Never"
  securityContext:
    fsGroup: 1000
  tolerations:
  - effect: "NoSchedule"
    key: "node.kubernetes.io/ci"
    operator: "Exists"
  - effect: "PreferNoSchedule"
    key: "node.kubernetes.io/ci"
    operator: "Exists"
  volumes:
  - hostPath:
      path: "/var/run/docker.sock"
    name: "volume-0"
  - hostPath:
      path: "/var/data/jenkins_sonar_cache"
    name: "volume-2"
  - hostPath:
      path: "/var/data/jenkins_go_cache"
    name: "volume-1"
  - emptyDir:
      medium: ""
    name: "workspace-volume"

Running on go-j8gf2 in /home/jenkins/agent/workspace/demo
[Pipeline] {
[Pipeline] stage
[Pipeline] { (ss)
[Pipeline] container
[Pipeline] {
[Pipeline] sh
+ go version
go version go1.12.10 linux/amd64
[Pipeline] }
[Pipeline] // container
[Pipeline] }
[Pipeline] // stage
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
Finished: SUCCESS

Mike

Rick 我用你方法测试了下 3.10 的镜像问题依旧

换 2.1.0 后就可以了

Rick

下面是我的镜像的详细信息，帮忙提供下你那边的：

[root@i-q73jfmqj ~]# docker pull kubesphere/builder-go:v3.1.0
v3.1.0: Pulling from kubesphere/builder-go
Digest: sha256:5b017039ba045c9d30904e97e65fed1239b3e87fb8519e383a25faa42bc2fd9c
Status: Image is up to date for kubesphere/builder-go:v3.1.0
docker.io/kubesphere/builder-go:v3.1.0

Mike

最最最后的结论：根本原因还是官方这个3.1 的镜像太大了 1.43 G 节点还没下载完引起的手动提前在节点上下载好镜像
就可以了