Kubernetes进阶 -- 编写资源清单（二）

容器列表

[root@server2 manifest]# vim pod.yml
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
    - name: myappv2
      image: myapp:v2		#在这里定义两个容器

[root@server2 manifest]# kubectl create -f pod.yml 
pod/myapp created

[root@server2 manifest]# kubectl get pod
NAME    READY   STATUS   RESTARTS   AGE
myapp   1/2     Error    4          2m5s      

[root@server2 manifest]# kubectl logs myapp myappv2
2020/06/26 02:57:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use)		# 这里
nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use)

由于myapp这个容器开起的是一个服务，所以在pod当中开启多个同样容器时，端口就会出现抢夺，其中一个容器就会一直重启，所以不能同时开启端口一样的两个容器。

# 修改：
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
    - name: demo
      image: busyboxplus		# 这是一个交互式的终端,需要下面三个参数的支持
      tty: true			# 打开一个终端，需要 stdin 和stdinOnce的支持。
      stdin: true		
      stdinOnce: true
      
[root@server2 manifest]# kubectl delete -f pod.yml 
pod "myapp" deleted
[root@server2 manifest]# kubectl apply -f pod.yml 
pod/myapp created
[root@server2 manifest]# kubectl get pod
NAME    READY   STATUS    RESTARTS   AGE
myapp   2/2     Running   0          10s

[root@server2 manifest]# kubectl attach myapp -c demo -it		#-c指定容器，-it 交互
If you don't see a command prompt, try pressing enter.
/ # 
/ # curl localhost
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

再这个容器却可以本地访问myapp：v1容器开启的端口，这时因为同一个pod中的容器共用网络。

那末对于myapp打开一个服务并没有打开一个终端的容器怎么连接哪？

[root@server2 manifest]# kubectl exec -it myapp -c myappv1 sh	#exec让他打开一个shell
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl kubectl exec [POD] -- [COMMAND] instead.
/ # netstat -tnlp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name    
tcp        0      0 0.0.0.0:80              0.0.0.0:*               LISTEN      1/nginx: master pro
/ # cat /etc/hostname 
myapp
/ # cd /usr/share/nginx/html/
/usr/share/nginx/html # cat index.html 
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

关闭这些服务后，k8s会自动进行重启。

镜像拉取策略

[root@server2 manifest]# kubectl explain pod.spec.containers
   imagePullPolicy	<string>
     Image pull policy. One of Always, Never, IfNotPresent. Defaults to Always
     if :latest tag is specified, or IfNotPresent otherwise. Cannot be updated.

默认总会去仓库拉取最新的，这就要求我们再更新镜像的时候打上一个最新的标签。

使用：

spec:
  containers:
    - name: myappv1
      image: myapp:v1
    - name: demo
      image: busyboxplus
      tty: true
      stdin: true
      stdinOnce: true
      imagePullPolicy: IfNotPresent
在这里加上参数就行了。IfNotPresent 当我们本地没有的时候才去拉取

端口映射

[root@server2 manifest]# vim pod.yml
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
      ports:
      - name: http
        containerPort: 80		#端口映射，就不用再命令行中加参数了
        hostPort: 80

[root@server2 manifest]# kubectl apply -f pod.yml 
pod/myapp created
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   0          6s    10.244.2.22   server4   <none>           <none>

# 再server4上查看端口映射
[root@server4 ~]# iptables -t nat -nL| grep :80
CNI-HOSTPORT-SETMARK  tcp  --  10.244.2.22          0.0.0.0/0            tcp dpt:80
CNI-HOSTPORT-SETMARK  tcp  --  127.0.0.1            0.0.0.0/0            tcp dpt:80
DNAT       tcp  --  0.0.0.0/0            0.0.0.0/0            tcp dpt:80 to:10.244.2.22:80

使用主机网络模式

[root@server2 manifest]# vim pod.yml 
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
  hostNetwork: true		#默认是false
#    - name: demo
#      image: busyboxplus
#      tty: true
#      stdin: true
#      stdinOnce: true
#      imagePullPolicy: IfNotPresent
[root@server2 manifest]# kubectl apply -f pod.yml 
kubepod/myapp created
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP             NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   0          5s    172.25.254.4   server4   <none>           <none>
[root@server2 manifest]# curl 172.25.254.4
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

172.25.254.4 正是我们server4的ip地址，共用了网络，但是这种容器只能开启一个，不能有副本，因为ip只有一个，会冲突。

pod重启策略

restartPolicy，定义pod的重启策略，有always，onfailure，never，默认为always，就是无论容器如何终止，时中去重启它，onfailure是只有pod以非0退出码终止时，才会重启它，就是当正常退出时（0）不重启，never，将退出码报告给master，永不重启。

[root@server2 manifest]# kubectl exec -it myapp -c myappv1 sh	# 交互式访问，
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl kubectl exec [POD] -- [COMMAND] instead.
/usr/share/nginx # nginx -s stop		# 关闭nginx
2020/06/26 03:58:21 [notice] 18#18: signal process started
/usr/share/nginx # command terminated with exit code 137
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   1          68s   10.244.2.23   server4   <none>           <none>
						## 进行了重启

现在我们加上never参数：

[root@server2 manifest]# kubectl delete -f pod.yml 		# 删除之前的
vim podpod "myapp" deleted

[root@server2 manifest]# vim pod.yml 
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
  restartPolicy: Never		# 永不重启

[root@server2 manifest]# kubectl apply -f pod.yml 
pod/myapp created
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   0          4s    10.244.2.24   server4   <none>           <none>
[root@server2 manifest]# kubectl exec -it myapp -c myappv1 sh
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl kubectl exec [POD] -- [COMMAND] instead.
/ # nginx -s stop
2020/06/26 04:01:52 [notice] 12#12: signal process started
/ # command terminated with exit code 137
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS      RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   0/1     Completed   0          20s   10.244.2.24   server4   <none>           <none>
		#容器就显示完成了，就不再重启

[root@server2 manifest]# kubectl get pod -o yaml |less
我们可以导出到yaml文件格式来查看具体的格式，这种方法还可以用与忘记格式时使用

resource资源限制

[root@server2 manifest]# vim pod.yml
apiVersion: v1
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
      resources:
        requests:		# 下限
          memory: 100Mi
          cpu: 0.1
        limits:			# 上限
          memory: 200Mi
          cpu: 0.2

[root@server2 manifest]# kubectl apply -f pod.yml 
kubec	getpod/myapp created
[root@server2 manifest]# kubectl describe pod myapp 
Name:         myapp
Namespace:    default
...
    Limits:
      cpu:     200m
      memory:  200Mi
    Requests:
      cpu:        100m
      memory:     100Mi
...

结点选择器

nodeSelector，定义node的label过滤标签，以key：value格式指定。

[root@server2 manifest]# kubectl get node --show-labels 
NAME      STATUS   ROLES    AGE     VERSION   LABELS
...
server4   Ready    <none>   7d15h   v1.18.3   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,
kubernetes.io/hostname=server4,kubernetes.io/os=linux

[root@server2 manifest]# vim pod.yml
kind: Pod
metadata:
  name: myapp
#  namespace: default
#  labels:
spec:
  containers:
    - name: myappv1
      image: myapp:v1
      resources:
        requests:
          memory: 100Mi
          cpu: 0.1
        limits:
          memory: 200Mi
          cpu: 0.2

 - [ ] nodeSelector:
           kubernetes.io/hostname: server4		选择server4的标签

[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   0          4s    10.244.2.25   server4   <none>           <none>
是在server4上.

[root@server2 manifest]# vim pod.yml 
 - [ ] nodeSelector:
           kubernetes.io/hostname: server3		改为server3的标签
[root@server2 manifest]# kubectl apply -f pod.yml 
pod/myapp created
[root@server2 manifest]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE      NOMINATED NODE   READINESS GATES
myapp   1/1     Running   0          3s    10.244.1.48   server3   <none>           <none>
就运行在server3上了

运行命令

指定容器的启动命令。

apiVersion: v1
kind: Pod
metadata:
  name: pod-example
spec:
  containers:
  - name: ubuntu
    image: ubuntu:trusty
    command: ["echo"]
    args: ["Hello World"]		# 加上这两个参数就行了，也可以都写到command中，用，隔开。

打标签

[root@server2 manifest]# vim pod.yml 
apiVersion: v1
kind: Pod
metadata:
  name: myapp
  labels:
    wo: myapp
spec:
  containers:
    - name: myappv1
      image: myapp:v1

[root@server2 manifest]# kubectl apply -f pod.yml 
pod/myapp created
[root@server2 manifest]# kubectl get pod --show-labels 
NAME    READY   STATUS    RESTARTS   AGE   LABELS
myapp   1/1     Running   0          14s   wo=myapp
[root@server2 manifest]# kubectl get pod -l wo		# 过滤包含wo的标签
NAME    READY   STATUS    RESTARTS   AGE
myapp   1/1     Running   0          90s
[root@server2 manifest]# kubectl get pod -L wo
NAME    READY   STATUS    RESTARTS   AGE   WO
myapp   1/1     Running   0          2m    myapp
[root@server2 manifest]# kubectl label pod myapp version=v1		#打新的标签
pod/myapp labeled
[root@server2 manifest]# kubectl get pod --show-labels 
NAME    READY   STATUS    RESTARTS   AGE     LABELS
myapp   1/1     Running   0          4m39s   version=v1,wo=myapp

[root@server2 manifest]# kubectl label pod myapp version=v2 --overwrite 		#更改标签
pod/myapp labeled
[root@server2 manifest]# kubectl get pod --show-labels 
NAME    READY   STATUS    RESTARTS   AGE     LABELS
myapp   1/1     Running   0          5m56s   version=v2,wo=myapp
[root@server2 manifest]# kubectl label pod myapp wo=ni --overwrite 
pod/myapp labeled
[root@server2 manifest]# kubectl get pod --show-labels 
NAME    READY   STATUS    RESTARTS   AGE    LABELS
myapp   1/1     Running   0          6m8s   version=v2,wo=ni