job配置和简易测试

1.job简单介绍

job也是种控制器,k8s有两种类型的控制器,一种是服务类控制器,比如deployment,deamonset,replicaset等等.一种是工作任务类控制器,job和cronjon就是工作任务类控制器.

job的简易参数介绍

spec.template格式同Pod

RestartPolicy仅支持Never或OnFailure

单个Pod时，默认Pod成功运行后Job即结束.

spec.completions标志Job结束需要成功运行的Pod个数,默认为1,可以理解pod运行的总数.

spec.parallelism标志并行运行的Pod的个数,默认为1,可以理解为同时运行的Pod数量.

spec.activeDeadlineSeconds标志失败Pod的重试最大时间,超过这个时间不会继续重试.

spec.backoffLimit: 指定job失败后进行重试的次数.

2.job简易实例

[root@k8s-node1 job]#cat job1.yaml

apiVersion: batch/v1

kind: Job

metadata:

name: pi

spec:

template:

metadata:

name: pi

spec:

containers:

- name: pi

image: perl

command: ["perl", "-Mbignum=bpi", "-wle", "print bpi(2000)"]

restartPolicy: Never

[root@k8s-node1 job]# kubectl get job

NAME COMPLETIONS DURATION AGE

pi 1/1 8m37s 15m

[root@k8s-node1 job]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 76 28d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 7 22d

mysql-t-54666b579c-7m5rv 1/1 Running 11 26d

mysql-test-647b8db96b-qdxw6 1/1 Running 10 23d

pi-5gtpz 0/1 Completed 0 15m

wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 7 22d

[root@k8s-node1 job]# kubectl logs pi-5gtpz

3.1415926535897932384626433832795028841971693993751058209749445923078164062862089986280348253421170679821480865132823066470938446095505822317253594081284811174502841027019385211055596446229489549303819644288109756659334461284756482337867831652712019091456485669234603486104543266482133936072602491412737245870066063155881748815209209628292540917153643678925903600113305305488204665213841469519415116094330572703657595919530921861173819326117931051185480744623799627495673518857527248912279381830119491298336733624406566430860213949463952247371907021798609437027705392171762931767523846748184676694051320005681271452635608277857713427577896091736371787214684409012249534301465495853710507922796892589235420199561121290219608640344181598136297747713099605187072113499999983729780499510597317328160963185950244594553469083026425223082533446850352619311881710100031378387528865875332083814206171776691473035982534904287554687311595628638823537875937519577818577805321712268066130019278766111959092164201989380952572010654858632788659361533818279682303019520353018529689957736225994138912497217752834791315155748572424541506959508295331168617278558890750983817546374649393192550604009277016711390098488240128583616035637076601047101819429555961989467678374494482553797747268471040475346462080466842590694912933136770289891521047521620569660240580381501935112533824300355876402474964732639141992726042699227967823547816360093417216412199245863150302861829745557067498385054945885869269956909272107975093029553211653449872027559602364806654991198818347977535663698074265425278625518184175746728909777727938000816470600161452491921732172147723501414419735685481613611573525521334757418494684385233239073941433345477624168625189835694855620992192221842725502542568876717904946016534668049886272327917860857843838279679766814541009538837863609506800642251252051173929848960841284886269456042419652850222106611863067442786220391949450471237137869609563643719172874677646575739624138908658326459958133904780275901

[root@k8s-node1 job]#

修改配置文件,加上参数测试

[root@k8s-node1 job]# cat job1.yaml

apiVersion: batch/v1

kind: Job

metadata:

name: pi

spec:

completions: 3

parallelism: 2

template:

metadata:

name: pi

spec:

containers:

- name: pi

image: perl

command: ["perl", "-Mbignum=bpi", "-wle", "print bpi(2000)"]

restartPolicy: Never

spec.parallelism参数设置了2,因此开始就启动了两个pod

[root@k8s-node1 job]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 76 28d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 7 22d

mysql-t-54666b579c-7m5rv 1/1 Running 11 26d

mysql-test-647b8db96b-qdxw6 1/1 Running 10 23d

pi-jjpm4 0/1 ContainerCreating 0 4s

pi-xffsz 0/1 ContainerCreating 0 4s

wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 7 22d

[root@k8s-node1 job]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 76 28d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 7 22d

mysql-t-54666b579c-7m5rv 1/1 Running 11 26d

mysql-test-647b8db96b-qdxw6 1/1 Running 10 23d

pi-c4qxs 0/1 Completed 0 43s

pi-jjpm4 0/1 Completed 0 55s

pi-xffsz 0/1 Completed 0 55s

wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 7 22d

3.job备份mysql容器数据库

job在什么场景使用呢?

比如数据库备份

最简单的示例,参考见下:

思路是启用一个job,使用的镜像是mysql:5.7(注意密码必须设置),然后执行命令mysqldump --host=mysql-test -uroot -ppassword --databases t1 >t1.sql.备份MySQL容器pod主机名为mysql-test的数据库(注意mysql-test是svc的名字).

[root@k8s-node1 job]# kubectl get svc,pod

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

service/httpd-svc NodePort 10.254.125.1 80:8400/TCP 31d

service/kubernetes ClusterIP 10.254.0.1 443/TCP 45d

service/mysql ClusterIP 10.254.209.23 3306/TCP 23d

service/mysql-t ClusterIP 10.254.177.63 3306/TCP 27d

service/mysql-test ClusterIP 10.254.177.188 3306/TCP 24d

service/wordpress NodePort 10.254.88.118 8080:8425/TCP 23d

NAME READY STATUS RESTARTS AGE

pod/busybox 1/1 Running 92 29d

pod/mysql-7c9cbfcdf8-hxlrm 1/1 Running 17 23d

pod/mysql-dump-jw5zh 0/1 Completed 0 80s

pod/mysql-t-54666b579c-7m5rv 1/1 Running 21 27d

pod/mysql-test-647b8db96b-qdxw6 1/1 Running 19 24d

pod/t1-55f6c78557-6xxwd 1/1 Running 0 17m

pod/t2-7f459d454c-wk2zb 1/1 Running 0 17m

pod/wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 12 23d

[root@k8s-node1 job]# cat job2.yaml

apiVersion: batch/v1

kind: Job

metadata:

name: mysql-dump

spec:

template:

metadata:

name: mysql-dump

spec:

containers:

- name: mysql-dump

image: mysql:5.7

env:

- name: MYSQL_ROOT_PASSWORD

value: abc123

command: ["/bin/sh","-c","mysqldump --host=mysql-test -uroot -ppassword --databases t1 >t1.sql"]

restartPolicy: Never

测试执行成功.见下:

[root@k8s-node1 storage]# kubectl get job

NAME COMPLETIONS DURATION AGE

mysql-dump 1/1 2s 28m

[root@k8s-node1 storage]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 81 29d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 7 22d

mysql-dump-p4wn7 0/1 Completed 0 28m

mysql-t-54666b579c-7m5rv 1/1 Running 11 27d

mysql-test-647b8db96b-qdxw6 1/1 Running 10 23d

nfs-client-provisioner-5bd47b7669-lzv6j 0/1 ContainerCreating 0 6m38s

wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 7 22d

完美点的示例,把备份的数据解耦出来,加上pvc.

配置pv和pvc,参考见下:

pv

[root@k8s-node1 storage]# cat job-pv1.yaml

apiVersion: v1

kind: PersistentVolume

metadata:

name: jobpv1

spec:

capacity:

storage: 100Mi

accessModes:

- ReadWriteMany

persistentVolumeReclaimPolicy: Retain

storageClassName: nfs

nfs:

path: /mnt/data/jobpv1

server: 192.168.174.130

[root@k8s-node1 storage]# kubectl apply -f job-pv1.yaml

persistentvolume/jobpv1 created

[root@k8s-node1 storage]# kubectl get pv

NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE

jobpv1 100Mi RWX Retain Available nfs 7s

pvc

[root@k8s-node1 storage]# cat job-pvc1.yaml

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

name: jobpvc1

spec:

accessModes:

- ReadWriteMany

resources:

requests:

storage: 100Mi

storageClassName: nfs

[root@k8s-node1 storage]# kubectl apply -f job-pvc1.yaml

persistentvolumeclaim/jobpvc1 created

[root@k8s-node1 storage]# kubectl get pvc

NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE

jobpvc1 Bound jobpv1 100Mi RWX nfs 4s

把pvc使用到job上

[root@k8s-node1 job]# cat job2.yaml

apiVersion: batch/v1

kind: Job

metadata:

name: mysql-dump

spec:

template:

metadata:

name: mysql-dump

spec:

containers:

- name: mysql-dump

image: mysql:5.7

env:

- name: MYSQL_ROOT_PASSWORD

value: abc123

command: ["/bin/sh","-c","mysqldump --host=mysql-test -uroot -ppassword --databases t1 >/mnt/t1.sql"]

volumeMounts:

- name: job-pvc

mountPath: "/mnt"

restartPolicy: Never

volumes:

- name: job-pvc

persistentVolumeClaim:

claimName: jobpvc1

[root@k8s-node1 job]# kubectl apply -f job2.yaml

job.batch/mysql-dump created

[root@k8s-node1 job]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 92 29d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 17 23d

mysql-dump-jw5zh 0/1 Completed 0 3s

mysql-t-54666b579c-7m5rv 1/1 Running 21 27d

mysql-test-647b8db96b-qdxw6 1/1 Running 19 24d

t1-55f6c78557-6xxwd 1/1 Running 0 16m

t2-7f459d454c-wk2zb 1/1 Running 0 15m

wordpress-pod-74c47cd8dd-dlzvc 1/1 Running 12 23d

可以去保存pvc的目录下查看,数据已经备份成功,见下

[root@k8s-node3 jobpv1]# pwd

/mnt/data/jobpv1

[root@k8s-node3 jobpv1]# ls

t1.sql

4.使用cronjob备份

上面用job备份只是一次性执行结束,对于很多备份是需要持续性备份操作的,持续性备份操作,k8s有cronjob.

就用上面的示例修改

注意cronjob的api和job的api不一样,可用expalin命令检索,见下

[root@k8s-node1 job]# kubectl explain cronjob

KIND: CronJob

VERSION: batch/v1beta1

修改好的cronjob示例见下

[root@k8s-node1 job]# cat cronjob1.yaml

apiVersion: batch/v1beta1

kind: CronJob

metadata:

name: mysql-dump

spec:

schedule: "*/1 * * * *"

jobTemplate:

spec:

template:

spec:

containers:

- name: mysql-dump

image: mysql:5.7

env:

- name: MYSQL_ROOT_PASSWORD

value: abc123

command: ["/bin/sh","-c","mysqldump --host=mysql-test -uroot -ppassword --databases t1 >/mnt/t1`date +%Y%m%d%H%M`.sql"]

volumeMounts:

- name: job-pvc

mountPath: "/mnt"

restartPolicy: Never

volumes:

- name: job-pvc

persistentVolumeClaim:

claimName: jobpvc1

注释1:schedule: "/1 ",每分钟执行一次注释2:date +%Y%m%d%H%M,date函数命令,文件名加上日期时间,更能直观*

读取执行cronjob

[root@k8s-node1 job]# kubectl apply -f cronjob1.yaml

cronjob.batch/mysql-dump created

[root@k8s-node1 job]# kubectl get cronjob

NAME SCHEDULE SUSPEND ACTIVE LAST SCHEDULE AGE

mysql-dump */1 * * * * False 1 5s 10s

[root@k8s-node1 job]# kubectl get pod

NAME READY STATUS RESTARTS AGE

busybox 1/1 Running 103 32d

mysql-7c9cbfcdf8-hxlrm 1/1 Running 21 26d

mysql-dump-1577082660-d58mx 0/1 Completed 0 9s

检索pvc存储目录,可见每分钟保存一个sql文件,见下:

[root@k8s-node3 jobpv1]# pwd

/mnt/data/jobpv1

[root@k8s-node3 jobpv1]# ll

total 32

-rw-r--r-- 1 root root 2447 Dec 23 14:31 t1201912230631.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:32 t1201912230632.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:33 t1201912230633.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:34 t1201912230634.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:35 t1201912230635.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:36 t1201912230636.sql

-rw-r--r-- 1 root root 2447 Dec 23 14:37 t1201912230637.sql

cronjob用来备份数据,非常方便.