返回总目录页

自动化部署、通知、分布式构建与并行构建(jenkins pipeline)

jenkins与ansible集成

 简单案例

pipeline {
    agent any
    stages {
        stage('deploy') {
            steps {
                ansiblePlaybook(
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant'
                )
            }
        }
    }
}

 

 

应该需要在Jenkins上安装一下

[root@mcw15 plugins]# yum install -y ansible

打开配置

[root@mcw15 plugins]# vim /etc/ansible/ansible.cfg 
[root@mcw15 plugins]# grep host_key_checking /etc/ansible/ansible.cfg
host_key_checking = False
[root@mcw15 plugins]#

也可以添加工具，添加不同的ansible版本，然后不同的项目用不同的ansible

 

不通：

[root@mcw15 mcwansible]# 
[root@mcw15 mcwansible]# ansible -i hosts example -m ping
10.0.0.13 | UNREACHABLE! => {
    "changed": false, 
    "msg": "Failed to connect to the host via ssh: Permission denied (publickey,password).", 
    "unreachable": true
}
[root@mcw15 mcwansible]#

分发公钥

[root@mcw15 mcwansible]#  ssh-copy-id -i /root/.ssh/id_rsa.pub 10.0.0.13
/usr/bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/root/.ssh/id_rsa.pub"
/usr/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed
/usr/bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys
root@10.0.0.13's password: 

Number of key(s) added: 1

Now try logging into the machine, with:   "ssh '10.0.0.13'"
and check to make sure that only the key(s) you wanted were added.

[root@mcw15 mcwansible]# ls
fenfa.sh  hosts
[root@mcw15 mcwansible]# cat hosts 
[example]
10.0.0.13
[root@mcw15 mcwansible]# ansible -i hosts example -m ping
10.0.0.13 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}
[root@mcw15 mcwansible]#

example2的先不让通

[root@mcw15 mcwansible]# vim hosts 
[root@mcw15 mcwansible]# cat hosts 
[example]
10.0.0.13
[example2]
10.0.0.22
[root@mcw15 mcwansible]# ansible -i hosts example2 -m ping
10.0.0.22 | UNREACHABLE! => {
    "changed": false, 
    "msg": "Failed to connect to the host via ssh: Warning: Permanently added '10.0.0.22' (ECDSA) to the list of known hosts.\r\nPermission denied (publickey,password).", 
    "unreachable": true
}
[root@mcw15 mcwansible]#

加个root作为凭证

 

暂时用这个项目做pipeline吧

 新增两个文件

hosts

[example]
10.0.0.13
[example2]
10.0.0.22

playbook.yml

- hosts: example2
  tasks:
    - debug: msg="{{ lookup('env','BUILD_TAG') }}"

 Jenkins file

pipeline {
    agent any
    stages {
        stage('deploy') {
            steps {
                ansiblePlaybook(
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant'
                )
            }
        }
    }
}

手动触发构建，失败

 

没这个步骤,步骤列表可以看到没有这个，现在有了docker了，之前安装了插件之后

java.lang.NoSuchMethodError: No such DSL method 'ansiblePlaybook' found among steps [

 安装ansible插件：https://plugins.jenkins.io/ansible

安装之后，重新构建。报错，权限。可能是我们在Jenkins上加的凭证vagrant没有起到效果吧

 把2去掉，用example，这个已经用密钥做了免密登录的

 

构建结果一样

 此时在机器上是正常执行命令的

 好吧，凭证填写了但是没有确定

 新增之后再次触发

 这次可以看到，已经成功了

 并且从在yml中用lookup获取到了Jenkins任务运行的环境变量

 此时改为example2，这个机器和Jenkins本身没有免密，执行命令的时候也会出现权限问题。看这里通过Jenkins凭证，是否可以正常执行

 可以看到，example2的也正常可以运行剧本

 另外的案例

Jenkinsfile

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    limit: 'example,example2',
                    tags: 'debugtag,testtag',
                    extraVars: [
                        login: 'mylogin',
                        secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
}

host

[example]
10.0.0.13
[example2]
10.0.0.22

playbook.yml

- hosts: example2
  tasks:
    - debug: msg="{{ lookup('env','BUILD_TAG') }}"

 手动触发构建，任务执行成功

 不过查看，只执行了example2,没有执行example的，不过这符合正常，因为我们在剧本里写死了是example2的主机运行

 ansibleVault步骤

不同资产不同的ansible登录方式

 指定这个机器 用这个machangwei用户和密码

 10.0.0.22

ansible_ssh_user: machangwei
ansible_become_method: sudo
ansible_ssh_pass: 123456

 hosts

[example]
10.0.0.13
[example2]
10.0.0.22

playbook.yml

- hosts: example2
  tasks:
    - debug: msg="{{ lookup('env','BUILD_TAG') }}"

jenkinsfile,

检查的时候用Jenkins凭证，但是执行的时候用的是host_vars里面的machangwei用户了，如果密码有问题，那么执行的时候会报错了

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    // credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    // limit: 'example,example2',
                    // tags: 'debugtag,testtag',
                    // extraVars: [
                    //     login: 'mylogin',
                    //     secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    // ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
}

点击构建，查看检查的时候，是Jenkins凭证，用的root检查的语法。然后执行的时候，没有指定用户，也就是用的host_vars里面的machangwei用户成功运行了，sudo运行的，并且两个机器上都有machangwei系统用户，都是有sudo权限的。

 我们这里想要验证下，的确是用的host_vars里面的machangwei用户，可以将里面的密码从123456改为1234567，如果构建数字54失败，报权限错误，那么的确是符合预期的。也就是我们可以根据不同的主机，进行不同的登录方式

 结果就是我们想的那样，报错无效的密码，权限拒绝了。

 通知

邮箱配置和邮件通知

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    // credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    // limit: 'example,example2',
                    // tags: 'debugtag,testtag',
                    // extraVars: [
                    //     login: 'mylogin',
                    //     secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    // ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
    post {
        failure {
            mail body: 'failure body', from: '1xxx32@163.com',  to: '8xx15@qq.com',subject: 'build status :('
        }
    }
}

 

 

报错，没有安装邮件的插件Mailer Plugin吗？

An internal error occurred during form field validation (HTTP 403). Please reload the page and if the problem persists, ask the administrator for help.
Jenkins将用这个地址发送通知邮件给项目拥有者.这里可以填写"foo@acme.org"或者像"Jenkins Daemon <foo@acme.org>"形式的内容.

 

 查看是安装了的

 

 这次没有报错了，按照上面的例子，

 

 先测试一下，发给下面的测试接收邮箱

 提示发送成功，但是却没看到发送以及接收邮件，先不管了，先保存一下

 修改添加邮件发送

 虽然部署阶段失败，但是post发送邮件这里显示成功的

 查看日志输出，执行了post了。但是并没有看到邮件的发出和接收

 重新做下系统配置，配置邮箱

注意，之前没有填写smtp服务器，是不行的。得填写，并且密码不是平时登录邮箱的密码，而是授权码。
授权码是用于登录第三方邮件客户端的专用密码。

 上面再次点击测试，可以看到已经发送并接收到了邮件

 此时再次点击构建，部署失败，就会执行post里失败发送邮件的部分

 失败日志

 对比pipeline配置，查看邮件内容，发现没有发送邮件，检查了下，刚刚改造邮箱的系统配置后，没有保存。保存一下，重新点击立即构建，然后可以看到发送构建失败的邮件了

对比查看 

 使用Email Extension 插件发送通知

安装 Emai Extension 插件： https://plugins.jenkins.io/email-ext

 

 使用这个

 配置

 

 

 把之前配置的邮件通知的清空

 修改，然后立即构建

 报错：

加个凭证，我们上面没有加邮箱的账号密码

 失败，不知道咋弄，有时间再研究吧

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    // credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    // limit: 'example,example2',
                    // tags: 'debugtag,testtag',
                    // extraVars: [
                    //     login: 'mylogin',
                    //     secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    // ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
    post {
        failure {
            emailext body:
              """<p>EXECUTED: job <b>\'${env.JOB_NAME}:${env.BUILD_NUMBER}\'
              </b></p><p>View console output at "<a href="${env.BUILD_URL}">
              ${env.JOB_NAME}:${env.BUILD_NUMBER}</a>"</p>
                <p><i>(Build log is attached)</i></p>""",
                compressLog: true,
                attachLog: true,
                recipientProviders: [culprits(), developers(), requestor(),brokenBuildSuspects()],
                replyTo: '1xx2@163.com',
                subject: "Status: ${currentBuild.result?:'SUCCESS'} - Job \'${env.JOB_NAME}:${env.BUILD_NUMBER}\'",
                to: "13xx2@163.com"
        }
    }
}

 

 钉钉通知

参考：https://blog.csdn.net/qq_42259469/article/details/136813262

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    // credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    // limit: 'example,example2',
                    // tags: 'debugtag,testtag',
                    // extraVars: [
                    //     login: 'mylogin',
                    //     secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    // ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
    post {
        failure {
            dingtalk(
                robot: 'Jenkins-Dingding',
                type:'MARKDOWN',
                title: "success: ${JOB_NAME}",
                text: ["- 失败构建:${JOB_NAME}项目!\n- 持续时间:${currentBuild.durationString}\n- 任务:#${JOB_NAME}\n-  请注意: @庞xx,@史xx 需要检查构建失败原因"],
            )
        }
    }
}

 

 

我这里DingTalk版本2.4.9

钉钉插件：dingding-notifications

 

 

 配置一下

 

 

 加密用的是钉钉里的加签

 pipeline这里只需要改  robot: 'Jenkins-Dingding',这里，用Jenkins系统配置里的id

 构建失败

 执行失败钉钉通知

 发出通知

 通知消息，多加了一行让谁注意，只是文本，没有@通知的能力

 

HTTP请求通知

我们先写一个第三方系统

from flask import Flask, render_template, send_file, session,request
import json

app=Flask(__name__,template_folder='./')
app.debug=True

@app.route("/",methods=["GET","POST"])
def index():
    if request.method == "POST":
        print(request.headers)
        print(request.data)
        return 'ok'
    return render_template('index.html')



if __name__ == '__main__':
    app.run("0.0.0.0",9527)

开启并访问项目

 

 提交请求

 

 安装插件

 开启项目

 修改Jenkinsfile。我们这里第三方系统，没有认证，所以这里就忽略认证这一步了，也就是注释了authentication

pipeline {
    agent any
    stages {
        stage('Syntax check ansible playbook') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    credentialsId: 'vagrant',
                    extras: '--syntax-check'
                )
            }
        }

        stage('deploy') {
            steps {
                ansiblePlaybook(
                    disableHostKeyChecking: true,
                    playbook: "${env.WORKSPACE}/playbook.yml",
                    inventory: "${env.WORKSPACE}/hosts",
                    // credentialsId: 'vagrant',
                    // skippedTags: 'debugtag',
                    forks: 2,
                    // limit: 'example,example2',
                    // tags: 'debugtag,testtag',
                    // extraVars: [
                    //     login: 'mylogin',
                    //     secret_key: [value: 'g4dfKWENpeF6pY05', hidden: true]
                    // ]
                    // startAtTask: 'task4'
                )
            }
        }
    }
    post {
        failure {
            script{
                def response = httpRequest(url:'http://192.168.0.106:9527/',acceptType:"APPLICATION_JSON",
                contentType:"APPLICATION_JSON",
                httpMode: "POST",
                // authentication: "http_request",
                customHeaders:[
                    [name: "headername", value:"headerValue"],
                    [name: "token",value:"secret", maskValue: true]
                ],
                requestBody:"{'buildNumber':'${env.BUILD_NUMBER}'}",
                timeout: 5,
                validResponseCodes: "200:302")
            
                echo "${response.status}"
                echo "${response.content}"
            }
                
        }
    }
}

 

点击立即构建

 查看日志，pipeline里面已经开启了请求，并且一直卡在这里不动

 查看第三方系统的服务端，可以看到，已经接收并处理了pipeline发起的请求通知，并且包含了token和headername两个自定义请求头，以及我们发送过去的内容，也就是Jenkins的构建数字。

 从服务端的记录时间看，18分收到请求，但是过了快十分钟了，pipeline对第三方系统发起请求之后，这里卡住一直不往下进行，没有去结束pipeline了，这不知道啥情况

 

 因为上一个构建没有停止，这下一个构建，还在等待可用的执行器，也就是资源还没释放

 只是前两个一直没取消成功

 一开始卡住的时候没看到这个终止，现在出现 了，那就终止一下

 也可以把它从post里面改到stage里面

 执行ansible之前，就执行这里并卡柱了，这个构建数字是85

 查看服务端已经处理了请求

 

 分布式构建 与并行构建

通过JNLP协议增加agent

指定端口

 跟书中例子不同，协议少了很多项

 

 

 

 

 

 

 agent连接master的方式

 登录到Jenkinsagent机器。我们这里用mcw12,执行上面的增加agent的命令，也就是下载jar包并启动。我们先用unix并且不用密码的

连接之后，就卡在这里，是前台运行，后面可以改成后台运行

 再次回到Jenkins node列表，可以看到node1有信息了，应该从之前没信息的不可用，变为可用节点了

 crt c 终止前台运行

 查看node1情况，又是不可用节点了

 点进去看状态提示，连接中断

 查看日志，连接成功以及连接中断都有记录

 查看负载情况

 Jenkins agent机器上查看信息，以及日志

^C[root@mcw12 jenkinsagent]# ls
agent.jar  secret-file
[root@mcw12 jenkinsagent]# ls /app/
remoting
[root@mcw12 jenkinsagent]# ls /app/remoting/
jarCache  logs
[root@mcw12 jenkinsagent]# ls /app/remoting/jarCache/
0F  1D  34  7F  AD  B2  CE  DC  EF
[root@mcw12 jenkinsagent]# ls /app/remoting/logs/
remoting.log.0
[root@mcw12 jenkinsagent]# less /app/remoting/logs/remoting.log.0 
[root@mcw12 jenkinsagent]# 
[root@mcw12 jenkinsagent]# cat /app/remoting/logs/remoting.log.0
Aug 24, 2024 2:01:56 PM hudson.remoting.Launcher createEngine
INFO: Setting up agent: node1
Aug 24, 2024 2:01:56 PM hudson.remoting.Engine startEngine
INFO: Using Remoting version: 3198.v03a_401881f3e
Aug 24, 2024 2:01:56 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 24, 2024 2:01:56 PM hudson.remoting.Launcher$CuiListener status
INFO: Locating server among [http://10.0.0.25:8080/]
Aug 24, 2024 2:01:57 PM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver resolve
INFO: Remoting server accepts the following protocols: [JNLP4-connect, Ping]
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Agent discovery successful
  Agent address: 10.0.0.25
  Agent port:    9812
  Identity:      52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Handshaking
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Connecting to 10.0.0.25:9812
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Server reports protocol JNLP4-connect-proxy not supported, skipping
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Trying protocol: JNLP4-connect
Aug 24, 2024 2:01:57 PM org.jenkinsci.remoting.protocol.impl.BIONetworkLayer$Reader run
INFO: Waiting for ProtocolStack to start.
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Remote identity confirmed: 52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 24, 2024 2:01:57 PM hudson.remoting.Launcher$CuiListener status
INFO: Connected
[root@mcw12 jenkinsagent]#

然jar包在后台运行起来

[root@mcw12 jenkinsagent]# ls
agent.jar  secret-file
[root@mcw12 jenkinsagent]# java -jar agent.jar -url http://10.0.0.25:8080/ -secret @secret-file -name node1 -workDir "/app" &
[1] 4121
[root@mcw12 jenkinsagent]# Aug 24, 2024 2:09:39 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 24, 2024 2:09:39 PM org.jenkinsci.remoting.engine.WorkDirManager setupLogging
INFO: Both error and output logs will be printed to /app/remoting
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher createEngine
INFO: Setting up agent: node1
Aug 24, 2024 2:09:39 PM hudson.remoting.Engine startEngine
INFO: Using Remoting version: 3198.v03a_401881f3e
Aug 24, 2024 2:09:39 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Locating server among [http://10.0.0.25:8080/]
Aug 24, 2024 2:09:39 PM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver resolve
INFO: Remoting server accepts the following protocols: [JNLP4-connect, Ping]
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Agent discovery successful
  Agent address: 10.0.0.25
  Agent port:    9812
  Identity:      52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Handshaking
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Connecting to 10.0.0.25:9812
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Server reports protocol JNLP4-connect-proxy not supported, skipping
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Trying protocol: JNLP4-connect
Aug 24, 2024 2:09:39 PM org.jenkinsci.remoting.protocol.impl.BIONetworkLayer$Reader run
INFO: Waiting for ProtocolStack to start.
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Remote identity confirmed: 52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 24, 2024 2:09:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Connected

[root@mcw12 jenkinsagent]#

再次查看，node1状态已经正常

 已经连接

 

 脚本命令行，我们试试

 按照他给的例子执行命令，可以看到，的确是在node1 ，也就是mcw12上执行了命令，并返回了结果。

 日志

 一些系统信息

 也可以点击断开连接

复制现有node1,新增node2

 创建，然后不修改配置直接保存，

 

发现好像jdk版本低了

[root@mcw14 opt]# echo 2ebe228de99679c32e9a2a8ea459f450264b71454db97da0bdffbd6161e6e1f9 > secret-file
[root@mcw14 opt]# curl -sO http://10.0.0.25:8080/jnlpJars/agent.jar
[root@mcw14 opt]# java -version
openjdk version "1.8.0_372"
OpenJDK Runtime Environment (build 1.8.0_372-b07)
OpenJDK 64-Bit Server VM (build 25.372-b07, mixed mode)
[root@mcw14 opt]# mkdir /app
[root@mcw14 opt]# java -jar agent.jar -url http://10.0.0.25:8080/ -secret @secret-file -name node2 -workDir "/app"
Error: A JNI error has occurred, please check your installation and try again
Exception in thread "main" java.lang.UnsupportedClassVersionError: hudson/remoting/Launcher has been compiled by a more recent version of the Java Runtime (class file version 55.0), this version of the Java Runtime only recognizes class file versions up to 52.0
    at java.lang.ClassLoader.defineClass1(Native Method)
    at java.lang.ClassLoader.defineClass(ClassLoader.java:756)
    at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:142)
    at java.net.URLClassLoader.defineClass(URLClassLoader.java:473)
    at java.net.URLClassLoader.access$100(URLClassLoader.java:74)
    at java.net.URLClassLoader$1.run(URLClassLoader.java:369)
    at java.net.URLClassLoader$1.run(URLClassLoader.java:363)
    at java.security.AccessController.doPrivileged(Native Method)
    at java.net.URLClassLoader.findClass(URLClassLoader.java:362)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:418)
    at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:352)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:351)
    at sun.launcher.LauncherHelper.checkAndLoadMain(LauncherHelper.java:621)
[root@mcw14 opt]#

升级jdk版本和master的一样之后，再次执行，就可以加入到集群了。jdk1.8不行，是因为master是java17,所以node也得同样版本么，不清楚。

[root@mcw14 opt]# java -jar agent.jar -url http://10.0.0.25:8080/ -secret @secret-file -name node2 -workDir "/app" &
[1] 2304
[root@mcw14 opt]# Aug 25, 2024 12:05:35 AM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:05:35 AM org.jenkinsci.remoting.engine.WorkDirManager setupLogging
INFO: Both error and output logs will be printed to /app/remoting
Aug 25, 2024 12:05:35 AM hudson.remoting.Launcher createEngine
INFO: Setting up agent: node2
Aug 25, 2024 12:05:35 AM hudson.remoting.Engine startEngine
INFO: Using Remoting version: 3198.v03a_401881f3e
Aug 25, 2024 12:05:35 AM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Locating server among [http://10.0.0.25:8080/]
Aug 25, 2024 12:05:36 AM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver resolve
INFO: Remoting server accepts the following protocols: [JNLP4-connect, Ping]
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Agent discovery successful
  Agent address: 10.0.0.25
  Agent port:    9812
  Identity:      52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Handshaking
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Connecting to 10.0.0.25:9812
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Server reports protocol JNLP4-connect-proxy not supported, skipping
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Trying protocol: JNLP4-connect
Aug 25, 2024 12:05:36 AM org.jenkinsci.remoting.protocol.impl.BIONetworkLayer$Reader run
INFO: Waiting for ProtocolStack to start.
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Remote identity confirmed: 52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:05:36 AM hudson.remoting.Launcher$CuiListener status
INFO: Connected

[root@mcw14 opt]#

 测试断开连接：

 

 

 重新连接启用

 

 再过一会，可以看到，连接上了

已经恢复为正常可用节点

 

 通过swarm插件增加agent

swarm插件:   https://plugins.jenkins.io/swarm

swarm客户端：  https://repo.jenkins-ci.org/releases/org/jenkins-ci/plugins/swarm-client/3.9/swarm-client-3.9.jar

 

安装插件

 在装了jdk的机器上下载客户端

[root@mcw14 ~]# java -version
openjdk version "1.8.0_372"
OpenJDK Runtime Environment (build 1.8.0_372-b07)
OpenJDK 64-Bit Server VM (build 25.372-b07, mixed mode)
[root@mcw14 ~]# cd /opt/
[root@mcw14 opt]# ls
otp_src_23.3  otp_src_23.3.tar.gz  rabbitmq-server-3.8.16-1.el7.noarch.rpm  sonarqube  sonarqube-7.6.zip
[root@mcw14 opt]# wget https://repo.jenkins-ci.org/releases/org/jenkins-ci/plugins/swarm-client/3.9/swarm-client-3.9.jar


[root@mcw14 opt]# ls
otp_src_23.3  otp_src_23.3.tar.gz  rabbitmq-server-3.8.16-1.el7.noarch.rpm  sonarqube  sonarqube-7.6.zip  swarm-client-3.9.jar
[root@mcw14 opt]#

 启动jar包，添加node失败，后面有时间再看这种方式的吧，暂时没有解决

[root@mcw14 opt]# java -jar swarm-client-3.9.jar -username machangwei -password 123456 -master http://10.0.0.25:8080/jenkins  -name swarm-node
Aug 25, 2024 12:03:54 AM hudson.plugins.swarm.Client main
INFO: Client.main invoked with: [-username machangwei -password 123456 -master http://10.0.0.25:8080/jenkins -name swarm-node]
Aug 25, 2024 12:03:54 AM hudson.plugins.swarm.Client run
INFO: Discovering Jenkins master
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
Aug 25, 2024 12:03:55 AM hudson.plugins.swarm.SwarmClient discoverFromMasterUrl
SEVERE: Failed to fetch swarm information from Jenkins, plugin not installed?
Aug 25, 2024 12:03:55 AM hudson.plugins.swarm.Client run
SEVERE: RetryException occurred
hudson.plugins.swarm.RetryException: Failed to fetch swarm information from Jenkins, plugin not installed?
    at hudson.plugins.swarm.SwarmClient.discoverFromMasterUrl(SwarmClient.java:229)
    at hudson.plugins.swarm.Client.run(Client.java:146)
    at hudson.plugins.swarm.Client.main(Client.java:119)

Aug 25, 2024 12:03:55 AM hudson.plugins.swarm.Client run
INFO: Retrying in 10 seconds

 

agent部分详解

any

目前有三个节点，node1,node2都还没有运行过任务

 修改文件，点击立即构建

 构建成功

查看日志，运行在node2

 查看node2的工作目录，workspace也是新建的目录，

[root@mcw14 opt]# ls /app/
remoting  workspace
[root@mcw14 opt]# ls /app/remoting/
jarCache  logs
[root@mcw14 opt]# ls /app/workspace/
parameters-example  parameters-example@tmp
[root@mcw14 opt]#

此时查看下node2构建历史，没看到

 不过负载统计可以看到刚刚在猛增

 通过标签指定agent

给node2添加标签

 

 node1 只有一个标签，node2两个

 指定jdk8的标签，也就是node1或者node2

pipeline {
    agent {
        label 'jdk8'
    }
    stages {
        stage('Build') {
            steps {
                echo "machangwei build"
            }
        }
    }    
}

 

 点击构建，运行在node2上了

 反复执行，都是调度到node2,改一下，node1添加一个标签

 

 添加一个并且，

pipeline {
    agent {
        label 'jdk8 && machangwei'
    }
    stages {
        stage('Build') {
            steps {
                echo "machangwei build"
            }
        }
    }    
}

 

 这次可以看到，是调度到node1了，只要node1符合。

 改成agent下面用node,这样还可以用自定义工作目录

pipeline {
    agent {
        node {
            label 'jdk8 && machangwei'
            customWorkspace '/var/lib/custom'
        }
    }
    stages {
        stage('Build') {
            steps {
                echo "machangwei build"
            }
        }
    }    
}

构建执行，原本是/app目录下，后面又显示运行在custom这个不存在的目录

 可以看到，的确是把代码都拉取到自定义的目录下了，并且这个目录也是新建的

 如果两个地方都写了，那么两个地方都有运行的一部分内容

pipeline {
    agent {
        node {
            label 'jdk8 && machangwei'
            customWorkspace '/var/lib/custom'
        }
    }
    stages {
        stage('Build') {
            agent {
                label 'jdk8 && windows'
            }
            steps {
                echo "machangwei build"
            }
        }
    }    
}

 不分配节点

每个阶段都运行在指定的agent，那么最外层的那个agent就设置为none

pipeline {
    agent none
    stages {
        stage('Build') {
            agent {
                label 'jdk8 && windows'
            }
            steps {
                echo "machangwei build"
            }
        }

        stage('Deploy') {
            agent {
                label 'jdk8 && machangwei'
            }
            steps {
                echo "machangwei deploy"
            }
        }
    }    
}

带有windows标签的是node2,带有machangwei标签的是node1,如下,build在node2运行，deploy阶段在node1运行，符合我们pipeline的指定agent要求

 when指令的beforeAgent选项

下面的有点问题，测试没有符合预期

pipeline {
    agent none
    stages {
        stage('Build') {
            agent {
                label 'jdk8 && windows'
            }
            steps {
                echo "machangwei build"
            }
        }

        stage('Deploy') {
            agent {
                label 'jdk8 && machangwei'
            }
            when {
                beforeAgent true
                branch 'prod'
            }
            steps {
                echo "machangwei deploy"
            }
        }
    }    
}

 

 

我们把上面那节的jenkinsfile添加一个when判断，

 并且新增prod分支

 立即构建，因为目前默认拉取的是main分支，不符合when的条件 prod,分支是否是prod返回的不是true,所以这个阶段跳过没运行

 修改该项目的指定拉取分区为prod

 查看已经是拉取prod分支了，但是还是没有执行deploy这个阶段

 我们打印一下环境变量，可以看到虽然是prod分支，但是环境变量里面是origin/prod，所以跟pipeline里面写的prod没有匹配上，

 去掉环境变量的打印，并且将分支名称改为origin/prod

 发现结果还是一样，还是没有匹配上

 

 将构建任务交给docker

下面的有异常，实验未成功

pipeline {
    agent {
        docker {
            label 'docker'
            image 'busybox'
        }
    }
    stages {
        stage('Build') {
            steps {
                echo "machangwei build"
                sh "touch machangwei.txt"
            }
        }

    }    
}

 

新增一个节点，加上docker标签

 新增节点，直接将它加到jenkins master机器上运行

 添加节点

[root@mcw15 opt]# ls
containerd  nexus-3.71.0-06  nexus-3.71.0-06-unix.tar.gz  rakudo-pkg  sonatype-work  swarm-client-3.9.jar
[root@mcw15 opt]# ps -ef|grep java
jenkins   36131      1  0 Aug24 ?        00:05:33 /usr/bin/java -Djava.awt.headless=true -jar /usr/share/java/jenkins.war --webroot=%C/jenkins/war --httpPort=8080
root      86112  38851  0 12:01 pts/1    00:00:00 grep --color=auto java
[root@mcw15 opt]# echo 1c21371902ac3ad7b0a2497b066d80f5d90662737a4cf6df88073c6cec547c9d > secret-file
[root@mcw15 opt]# curl -sO http://10.0.0.25:8080/jnlpJars/agent.jar
[root@mcw15 opt]# mkdir /app
[root@mcw15 opt]# java -jar agent.jar -url http://10.0.0.25:8080/ -secret @secret-file -name "node-docker" -workDir "/app" &
[1] 86161
[root@mcw15 opt]# Aug 25, 2024 12:02:40 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:02:40 PM org.jenkinsci.remoting.engine.WorkDirManager setupLogging
INFO: Both error and output logs will be printed to /app/remoting
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher createEngine
INFO: Setting up agent: node-docker
Aug 25, 2024 12:02:40 PM hudson.remoting.Engine startEngine
INFO: Using Remoting version: 3198.v03a_401881f3e
Aug 25, 2024 12:02:40 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Locating server among [http://10.0.0.25:8080/]
Aug 25, 2024 12:02:40 PM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver resolve
INFO: Remoting server accepts the following protocols: [JNLP4-connect, Ping]
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Agent discovery successful
  Agent address: 10.0.0.25
  Agent port:    9812
  Identity:      52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Handshaking
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Connecting to 10.0.0.25:9812
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Server reports protocol JNLP4-connect-proxy not supported, skipping
Aug 25, 2024 12:02:40 PM hudson.remoting.Launcher$CuiListener status
INFO: Trying protocol: JNLP4-connect
Aug 25, 2024 12:02:40 PM org.jenkinsci.remoting.protocol.impl.BIONetworkLayer$Reader run
INFO: Waiting for ProtocolStack to start.
Aug 25, 2024 12:02:41 PM hudson.remoting.Launcher$CuiListener status
INFO: Remote identity confirmed: 52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:02:41 PM hudson.remoting.Launcher$CuiListener status
INFO: Connected

已经正常。这个应该一个服务器里面可以跑多个node程序的。这样应该可以提高并发处理能力。增加执行器。

 docker插件安装了的,不过不是我们用的那个

 下面这个才是：https://plugins.jenkins.io/docker-plugin，而不是下面那个workflow

安装好插件重启jenkins服务的时候，各个agent也是会被终止重连

[root@mcw14 opt]# Aug 25, 2024 12:16:15 PM hudson.remoting.Launcher$CuiListener status
INFO: Terminated
Aug 25, 2024 12:16:28 PM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver waitForReady
INFO: Controller isn't ready to talk to us on http://10.0.0.25:8080/tcpSlaveAgentListener/. Will try again: response code=503
Aug 25, 2024 12:16:38 PM hudson.remoting.Launcher$CuiListener status
INFO: Performing onReconnect operation.
Aug 25, 2024 12:16:39 PM jenkins.slaves.restarter.JnlpSlaveRestarterInstaller$EngineListenerAdapterImpl onReconnect
INFO: Restarting agent via jenkins.slaves.restarter.UnixSlaveRestarter@7a46cfb1
Aug 25, 2024 12:16:39 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:16:39 PM org.jenkinsci.remoting.engine.WorkDirManager setupLogging
INFO: Both error and output logs will be printed to /app/remoting
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher createEngine
INFO: Setting up agent: node2
Aug 25, 2024 12:16:39 PM hudson.remoting.Engine startEngine
INFO: Using Remoting version: 3198.v03a_401881f3e
Aug 25, 2024 12:16:39 PM org.jenkinsci.remoting.engine.WorkDirManager initializeWorkDir
INFO: Using /app/remoting as a remoting work directory
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Locating server among [http://10.0.0.25:8080/]
Aug 25, 2024 12:16:39 PM org.jenkinsci.remoting.engine.JnlpAgentEndpointResolver resolve
INFO: Remoting server accepts the following protocols: [JNLP4-connect, Ping]
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Agent discovery successful
  Agent address: 10.0.0.25
  Agent port:    9812
  Identity:      52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Handshaking
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Connecting to 10.0.0.25:9812
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Server reports protocol JNLP4-connect-proxy not supported, skipping
Aug 25, 2024 12:16:39 PM hudson.remoting.Launcher$CuiListener status
INFO: Trying protocol: JNLP4-connect
Aug 25, 2024 12:16:39 PM org.jenkinsci.remoting.protocol.impl.BIONetworkLayer$Reader run
INFO: Waiting for ProtocolStack to start.
Aug 25, 2024 12:16:41 PM hudson.remoting.Launcher$CuiListener status
INFO: Remote identity confirmed: 52:34:14:5f:aa:87:d8:d8:c9:a3:23:e7:32:15:da:d8
Aug 25, 2024 12:16:41 PM hudson.remoting.Launcher$CuiListener status
INFO: Connected

jenkins日志

 需要新增一个插件，并且docker-plugin插件需要降低版本，

 版本适配了，也不需要依赖包了

 使用这个之后，发现还是不行。可能是版本太低，不支持吧

WorkflowScript: 3: Invalid agent type "docker" specified. Must be one of [any, label, none] @ line 3, column 9.

 

 配置docker私有仓库

 并行构建

如下，分别在Chrome，Firefox，IE等浏览器的不同版本对用一个web应用进行UI测试，用下面方法，按照顺序执行，测试效率太慢了

pipeline {
    agent any
    stages {
        stage('Build') {
            steps {
                echo "Building..."
            }
        }
        stage('Test on Chrome') {
            steps {
                echo "Testing..."
            }
        }
        stage('Test on Firfox') {
            steps {
                echo "Testing..."
            }
        }

    }    
}

 

 

 修改，使用并行构建

 

 

pipeline {
    agent any
    stages {
        stage('Run Tests') {
            failFast true
            parallel {
                stage('Test On Chrome') {
                    agent { label "chrome" }
                    steps {
                        echo "Chrome UI测试"
                    }
                }
                stage('Test on Firfox') {
                    agent { label "firefox" }
                    steps {
                        echo "Firfox UI测试"
                    }
                }
                stage('Test on IE') {
                    agent { label "ie" }
                    steps {
                        echo "IE UI测试"
                    }
                }
            } //end of parallel 
        } // end of run test 
    } // end of stages
} // end of pipeline

 

 给三个agent上分别添加上上面对应的三个标签

 总共花了778ms，而之前的pipeline执行是花了959ms，这效率变高了

 从日志输出也可以看出，整个pipeline在master上运行，然后三个阶段分别在三个agent上运行，并且不是按照顺序完成一个，才开始下一个阶段的运行，而是并行。如果是顺序执行的话，那么running on 节点的日志输出不会放在一起出来，而是中间隔了好多行其它输出，也就是等一个阶段的所有输出结算之后，才会进行第二个阶段的日志输出。也就是这里是已经实现了并行，提高了效率。

 

 

 在不同的分支上应用并行构建

并行步骤

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

posted @ 2024-08-21 01:18  马昌伟  阅读(96)  评论(0)  编辑  收藏  举报
博主链接地址：https://www.cnblogs.com/machangwei-8/