一佳互联
虚拟化容器,大数据,DBA,中间件,监控。

k8s常用资源之rc资源（四）

11 11月
作者:admin|分类:容器虚拟化

k8s常用资源之rc资源

1.k8s资源常见操作

创建一个资源
kubectl create -f xxx.yaml

查看一个资源
kubectl get pod | rc

查看一个资源的详细过程
kubectl describe pos pod_name

删除一个资源
kubectl delete pod pod_name
kubectl delete -f xxx.yaml

编辑一个资源的配置文件
kubectl edit pod pod_name

2.rc资源

rc保证指定数量的pod始终存活，rc通过标签选择器来关联pod

rc创建的pod资源会均匀的分布在每个节点上，假如一个节点挂了，会立刻迁移到正常的节点上，始终保持pod资源存活，假如指定pod资源个数为5个，当挂了一个pod会立刻在创建一个新的pod，当多了一个pod会立马删除

删除多余pod资源时会删除存活时间最短的哪一个

rc在管理pod时通过 selector中定义的标签来管理pod的数量，如果pod标签与rc标签一致，rc则认为这是自己开启的pod会对它进行管理，当少的时候增加，多的时候删除

2.1.运行一个rc

1）辨析eyml文件
[root@k8s-master rc]# vim k8s_rc.yml 
apiVersion: v1				//api版本
kind: ReplicationController				//资源类型这里选择rc
metadata:					//属性
  name: nginxrc				//rc资源的名称
spec:
  replicas: 5				//启动几个pod资源
  selector:					//rc资源的标签
    app: myweb
  template:				//定义pod资源的目标
    metadata:	
      labels:
        app: myweb			//标签
    spec:
      containers:
      - name: myweb				//容器的名称
        image: 192.168.81.240/k8s/nginx:1.15			//容器仓库地址
        ports:
        - containerPort: 80			//端口

2）创建rc资源
[root@k8s-master rc]# kubectl create -f k8s_rc.yml
replicationcontroller "nginxrc" created

3）查看rc资源
[root@k8s-master rc]# kubectl get rc
NAME      DESIRED   CURRENT   READY     AGE
nginxrc   5         5         5         16m
DESIRED表示要创建几个pod
CURRENT表示当前起来几个
READY表示准备好了几个

4）查看启动的pod，一共有5个
[root@k8s-master rc]# kubectl get pod
NAME            READY     STATUS    RESTARTS   AGE
nginxrc-5nd7q   1/1       Running   0          23s
nginxrc-f0np6   1/1       Running   0          24s
nginxrc-n7pg4   1/1       Running   0          23s
nginxrc-q0zln   1/1       Running   0          24s
nginxrc-tvzvj   1/1       Running   0          24s

5）查看pod资源是否均匀
[root@k8s-master rc]# kubectl get pod -o wide
NAME            READY     STATUS    RESTARTS   AGE       IP            NODE
nginx2          1/1       Running   0          2d        172.16.77.2   192.168.81.230
nginxrc-5nd7q   1/1       Running   0          7m        172.16.46.3   192.168.81.220
nginxrc-f0np6   1/1       Running   0          7m        172.16.46.4   192.168.81.220
nginxrc-ldtnz   1/1       Running   0          35s       172.16.77.5   192.168.81.230
nginxrc-n7pg4   1/1       Running   0          7m        172.16.77.3   192.168.81.230
nginxrc-q0zln   1/1       Running   0          7m        172.16.77.4   192.168.81.230
test            2/2       Running   3          51m       172.16.46.2   192.168.81.220

2.2.删除一个rc运行的pod资源

1)删除
[root@k8s-master rc]# kubectl delete pod nginxrc-tvzvj
pod "nginxrc-tvzvj" deleted

2)删除完再查看，发现还是启动了5个，始终保证有5个
[root@k8s-master rc]# kubectl get pod
NAME            READY     STATUS    RESTARTS   AGE
nginxrc-5nd7q   1/1       Running   0          6m
nginxrc-f0np6   1/1       Running   0          6m
nginxrc-ldtnz   1/1       Running   0          9s
nginxrc-n7pg4   1/1       Running   0          6m
nginxrc-q0zln   1/1       Running   0          6m

2.3.node节点异常自动迁移

1)模拟node节点故障
[root@k8s-master rc]# kubectl delete node 192.168.81.230
node "192.168.81.230" deleted

2)查看pod资源是否迁移
[root@k8s-master rc]# kubectl get pod -o wide
NAME            READY     STATUS    RESTARTS   AGE       IP            NODE
nginxrc-27rlw   1/1       Running   0          1m        172.16.46.5   192.168.81.220
nginxrc-2scm1   1/1       Running   0          1m        172.16.46.7   192.168.81.220
nginxrc-5nd7q   1/1       Running   0          8m        172.16.46.3   192.168.81.220
nginxrc-f0np6   1/1       Running   0          8m        172.16.46.4   192.168.81.220
nginxrc-n3q8j   1/1       Running   0          1m        172.16.46.6   192.168.81.220

2.4.模拟增加pod资源观察删除的资源

1)修改test pod资源的标签
标签一致rc才会进行统一管理，始终保持指定的pod数量
将此pod的标签更换为nginxrc一样的标签
修改第十行
[root@k8s-master rc]# kubectl edit pod test 
app: myweb
pod "test" edited

2)观察
换之前的pod情况
[root@k8s-master rc]# kubectl get pod -o wide
NAME            READY     STATUS    RESTARTS   AGE       IP            NODE
nginxrc-27rlw   1/1       Running   0          1m        172.16.46.5   192.168.81.220
nginxrc-2scm1   1/1       Running   0          1m        172.16.46.7   192.168.81.220
nginxrc-5nd7q   1/1       Running   0          9m        172.16.46.3   192.168.81.220
nginxrc-f0np6   1/1       Running   0          9m        172.16.46.4   192.168.81.220
nginxrc-v8p5j   1/1       Running   0          1s        172.16.77.2   192.168.81.230
test            2/2       Running   3          53m       172.16.46.2   192.168.81.220

换之后的pod情况
[root@k8s-master rc]# kubectl get pod -o wide
NAME            READY     STATUS    RESTARTS   AGE       IP            NODE
nginxrc-27rlw   1/1       Running   0          2m        172.16.46.5   192.168.81.220
nginxrc-2scm1   1/1       Running   0          2m        172.16.46.7   192.168.81.220
nginxrc-5nd7q   1/1       Running   0          9m        172.16.46.3   192.168.81.220
nginxrc-f0np6   1/1       Running   0          9m        172.16.46.4   192.168.81.220
test            2/2       Running   3          54m       172.16.46.2   192.168.81.220

删除多余pod资源时会删除存活时间最短的那一个

在这里插入图片描述

2.5.rc一键升级和自动回滚

一键升级其实就是把之前的pod一个个停掉，然后在根据新的配置启动新pod

回滚就是把新版本退回到旧版本

语法配置：kubectl rolling-update nginxrc -f k8s_rc2.yml --update-period=10s

rolling-update：升级回滚都需要的关键字

nginxrc：表示对哪个资源进行升级或回滚

-f：指定yaml文件

–update-period=10s：默认是1分钟，我们用10s升级，1s回滚

2.5.1.将nginx1.13版本的rc资源升级为nginx1.15

1)查看旧版本的pod容器
[root@k8s-master rc]# kubectl get pod -o wide
NAME             READY     STATUS    RESTARTS   AGE       IP            NODE
nginxrc-ff9bc    1/1       Running   0          13m       172.16.16.2   192.168.81.230
nginxrc-h36p6    1/1       Running   0          13m       172.16.56.3   192.168.81.220
nginxrc-rqq14    1/1       Running   0          13m       172.16.16.3   192.168.81.230
nginxrc-ttqww    1/1       Running   0          13m       172.16.56.2   192.168.81.220
nginxrc-vqv6t    1/1       Running   0          13m       172.16.56.4   192.168.81.220
nginxrc2-cdvbf   1/1       Running   0          6s        172.16.16.4   192.168.81.230

[root@k8s-master rc]# curl -I 172.16.56.4 
HTTP/1.1 200 OK
Server: nginx/1.13.12

2)将nginx1.13升级为1.15
编写1.15版本的yaml
[root@k8s-master rc]# vim k8s_rc2.yml 
apiVersion: v1
kind: ReplicationController
metadata:
  name: nginxrc2
spec:
  replicas: 5
  selector:
    app: myweb2
  template:
    metadata:
      labels:
        app: myweb2
    spec:
      containers:
      - name: myweb2
        image: 192.168.81.240/k8s/nginx:1.15
        ports:
        - containerPort: 80

3)一键升级，启动一个新的停止一个旧的
[root@k8s-master rc]# kubectl rolling-update nginxrc -f k8s_rc2.yml  --update-period=10s
Continuing update with existing controller nginxrc2.
Scaling up nginxrc2 from 1 to 5, scaling down nginxrc from 5 to 0 (keep 5 pods available, don't exceed 6 pods)
Scaling nginxrc down to 4
Scaling nginxrc2 up to 2
Scaling nginxrc down to 3
Scaling nginxrc2 up to 3
Scaling nginxrc down to 2
Scaling nginxrc2 up to 4
Scaling nginxrc down to 1
Scaling nginxrc2 up to 5
Scaling nginxrc down to 0
Update succeeded. Deleting nginxrc
replicationcontroller "nginxrc" rolling updated to "nginxrc2

4)查看nginx版本
[root@k8s-master rc]# curl -I 172.16.56.3
HTTP/1.1 200 OK
Server: nginx/1.15.12

操作步骤详图

在这里插入图片描述

升级前后yaml文件比对

在这里插入图片描述

2.5.2.将nginx1.15版本回滚到nginx1.13

[root@k8s-master rc]# kubectl rolling-update nginxrc2 -f k8s_rc.yml --update-period=1s
Continuing update with existing controller nginxrc.
Scaling up nginxrc from 1 to 5, scaling down nginxrc2 from 5 to 0 (keep 5 pods available, don't exceed 6 pods)
Scaling nginxrc2 down to 4
Scaling nginxrc up to 2
Scaling nginxrc2 down to 3
Scaling nginxrc up to 3
Scaling nginxrc2 down to 2
Scaling nginxrc up to 4
Scaling nginxrc2 down to 1
Scaling nginxrc up to 5
Scaling nginxrc2 down to 0
Update succeeded. Deleting nginxrc2
replicationcontroller "nginxrc2" rolling updated to "nginxrc"
[root@k8s-master rc]# curl -I 172.16.16.2 2>/dev/null | head -3
HTTP/1.1 200 OK
Server: nginx/1.13.12
Date: Sun, 19 Jul 2020 16:14:55 GMT

在这里插入图片描述

更多文章推荐

历史上的今天
11月
11
浏览503 评论0
返回
目录 返回
首页
k8s常用资源之deployment资源（六） k8s常用资源之pod资源（三）
↑