Docker&Kubernetes ❀ Kubernetes集群Pod控制器 - Job

Docker&Kubernetes ❀ Kubernetes集群Pod控制器 - Job

文章目录

​​1、基本概念与控制器特点​​​​2、资源配置清单​​​​3、创建控制器​​

​​3.1 Job控制单个Pod​​​​3.2 Job控制多个并发Pod​​

​​4、删除控制器​​

1、基本概念与控制器特点

Job：主要用于负责批量处理（一次性处理指定数量任务）短暂的一次性（每个任务仅运行一次就结束）任务；

特点如下：

2、资源配置清单

参数查询方法：[root@master ~]# kubectl explain job

参数汇总梳理：

apiVersion: batch/v1kind: Jobmetadata: name: pc-job namespace: dev labels: controller: job spec: completions: 6 #指定job需要成功运行Pod的次数，默认值为1 parallelism: 3 #指定job在任一时刻应该并发运行pod的数量，默认值为1 activeDeadlineSeconds: 30 #指定job可运行的时间期限，超过时间未结束系统会尝试终止 backoffLimit: 6 #指定job失败后进行重试的次数，默认为6 manualSelector: true #是否可以使用selector选择器选择pod，默认为false selector: #pod选择器，指定控制那些pod matchLabels: app: count-pod matchExpressions: - key: "app" operator: In values: ["nginx-pod"] template: #模板，当副本数量不足时，会根据以下配置创建副本 metadata: labels: app: counter-pod spec: restartPolicy: Never #重启策略，支持Nerver或OnFailure containers: - name: counter image: busybox:latest command: ["/bin/sh","-c","for i in 9 8 7 6 5 4 3 2 1;do echo $i;sleep 2;done"]

重启策略：

如果指定为OnFailure，则Job会在Pod出现故障时重启容器，而不是创建Pod，failer次数不会改变；如果指定为Never，则Job会在Pod出现故障时创建新的Pod，并且故障Pod不会消失，failed次数加1；如果指定为Always，则一直重启，意味着Job任务会重复执行，因此不能设置为Always；

3、创建控制器

下面主要介绍两个使用，首先使用Job创建单个Pod以观察Pod创建的具体过程，其次创建多个Pod，使其进行并发创建，然后观察创建过程；

3.1 Job控制单个Pod

此实验建议开启三个终端，第一个终端执行调用文件命令，第二个终端监控Job控制器，第三个终端监控Pod；

#创建YAML文件[root@master ~]# cat pc-job.yaml apiVersion: batch/v1kind: Jobmetadata: name: pc-job namespace: devspec: manualSelector: true selector: matchLabels: app: counter-pod template: metadata: labels: app: counter-pod spec: restartPolicy: Never containers: - name: counter image: busybox:latest command: ["/bin/sh","-c","for i in 9 8 7 6 5 4 3 2 1;do echo $i;sleep 2;done"]

#调用YAML文件[root@master ~]# kubectl apply -f pc-job.yaml job.batch/pc-job created#监控Job控制器[root@master ~]# kubectl get job -n dev -w NAME COMPLETIONS DURATION AGEpc-job 0/1 0spc-job 0/1 0s 0spc-job 0/1 35s 35spc-job 1/1 35s 35s#监控PodNAME READY STATUS RESTARTS AGEpc-job-qtw84 0/1 Pending 0 0spc-job-qtw84 0/1 Pending 0 0spc-job-qtw84 0/1 ContainerCreating 0 0spc-job-qtw84 0/1 ContainerCreating 0 1spc-job-qtw84 1/1 Running 0 17spc-job-qtw84 0/1 Completed 0 35spc-job-qtw84 0/1 Completed 0 35spc-job-qtw84 0/1 Completed 0 35s

3.2 Job控制多个并发Pod

此实验仍然建议开启三个终端，第一个终端执行调用文件命令，第二个终端监控Job控制器，第三个终端监控Pod；

#创建YAML文件[root@master ~]# vim pc-job.yamapiVersion: batch/v1kind: Jobmetadata: name: pc-job namespace: devspec: manualSelector: true completions: 6 #指定job需要成功运行Pod的次数，默认值为1 parallelism: 3 #指定job在任一时刻应该并发运行pod的数量，默认值为1 selector: matchLabels: app: counter-pod template: metadata: labels: app: counter-pod spec: restartPolicy: Never containers: - name: counter image: busybox:latest command: ["/bin/sh","-c","for i in 9 8 7 6 5 4 3 2 1;do echo $i;sleep 2;done"]

#调用YAML文件[root@master ~]# kubectl apply -f pc-job.yaml job.batch/pc-job created#监控Job控制器[root@master ~]# kubectl get job -n dev -w NAME COMPLETIONS DURATION AGEpc-job 0/6 0spc-job 0/6 0s 0spc-job 0/6 35s 35spc-job 1/6 35s 35spc-job 1/6 36s 36spc-job 2/6 36s 36spc-job 2/6 52s 52spc-job 3/6 52s 52spc-job 3/6 71s 71spc-job 4/6 71s 71spc-job 4/6 71s 71spc-job 5/6 71s 71spc-job 5/6 87s 87spc-job 6/6 87s 87s#监控Pod[root@master ~]# kubectl get pod -n dev -wNAME READY STATUS RESTARTS AGEpc-job-jllsm 0/1 Pending 0 0spc-job-882tp 0/1 Pending 0 0spc-job-xblq4 0/1 Pending 0 0s#与配置参数符合，一次性创建3个Pod执行Job任务，共6个任务，创建2次pc-job-882tp 0/1 Pending 0 0spc-job-jllsm 0/1 Pending 0 0spc-job-xblq4 0/1 Pending 0 0spc-job-882tp 0/1 ContainerCreating 0 0spc-job-xblq4 0/1 ContainerCreating 0 0spc-job-jllsm 0/1 ContainerCreating 0 0spc-job-882tp 0/1 ContainerCreating 0 1spc-job-jllsm 0/1 ContainerCreating 0 2spc-job-xblq4 0/1 ContainerCreating 0 2spc-job-882tp 1/1 Running 0 17spc-job-jllsm 1/1 Running 0 18spc-job-xblq4 1/1 Running 0 34spc-job-882tp 0/1 Completed 0 35spc-job-8scd8 0/1 Pending 0 0spc-job-8scd8 0/1 Pending 0 0spc-job-882tp 0/1 Completed 0 35spc-job-8scd8 0/1 ContainerCreating 0 0spc-job-882tp 0/1 Completed 0 35spc-job-jllsm 0/1 Completed 0 36spc-job-n8hd9 0/1 Pending 0 0spc-job-n8hd9 0/1 Pending 0 0spc-job-n8hd9 0/1 ContainerCreating 0 0spc-job-jllsm 0/1 Completed 0 36spc-job-jllsm 0/1 Completed 0 36spc-job-8scd8 0/1 ContainerCreating 0 1spc-job-n8hd9 0/1 ContainerCreating 0 1spc-job-xblq4 0/1 Completed 0 52spc-job-zs97t 0/1 Pending 0 0spc-job-xblq4 0/1 Completed 0 52spc-job-zs97t 0/1 Pending 0 0spc-job-zs97t 0/1 ContainerCreating 0 0spc-job-xblq4 0/1 Completed 0 52spc-job-8scd8 1/1 Running 0 17spc-job-zs97t 0/1 ContainerCreating 0 1spc-job-n8hd9 1/1 Running 0 17spc-job-zs97t 1/1 Running 0 17spc-job-8scd8 0/1 Completed 0 36spc-job-8scd8 0/1 Completed 0 36spc-job-8scd8 0/1 Completed 0 36spc-job-n8hd9 0/1 Completed 0 35spc-job-n8hd9 0/1 Completed 0 35spc-job-n8hd9 0/1 Completed 0 35spc-job-zs97t 0/1 Completed 0 35spc-job-zs97t 0/1 Completed 0 35spc-job-zs97t 0/1 Completed 0 35s

4、删除控制器

[root@master ~]# kubectl delete -f pc-job.yaml job.batch "pc-job" deleted

版权声明：本文内容由网络用户投稿，版权归原作者所有，本站不拥有其著作权，亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容，请联系我们jiasou666@gmail.com 处理，核实后本网站将在24小时内删除侵权内容。