[K8S] Pod资源共享

谁践踏了优雅 2022-09-04 15:53 242阅读 0赞

**@Pod资源共享机制**

Kubernetes直接管理Pod，而不是容器  
Pod不是进程，而是容器运行的环境  
Pod是Kubernetes创建和管理的最小单元，不同Pod可以在不同Node上  
Pod中可以有一个或多个容器\[1\]，且始终部署在一个Node上，这些容器共享存储、网络

说明\[1\]:  
Pod中运行单个容器 Pod可以看作是单个容器的抽象封装  
Pod中运行多个容器 需要共享资源或紧密耦合的应用程序\[2\],Pod 将这些容器和存储资源打包为一个可管理的实体。

说明\[2\]:   
应用程序之间发生文件、数据交互  
应用程序之间需要通过本地通信(127.0.0.1、socket)  
应用程序之间频繁调用

**官方文档传送门 **  
Pods -> https://kubernetes.io/zh/docs/concepts/workloads/pods/  
Init容器 -> https://kubernetes.io/zh/docs/concepts/workloads/pods/init-containers

在每个Node上执行docker ps 会发现有不少pause

pause容器/infra (infrastucture container)容器和业务容器一一对应，业务容器共享pause容器的网络栈和Volume挂载卷。

[root@k8s-master ~]# docker ps | grep dash
    64cea4b64eb1   86262685d9ab                                        "/metrics-sidecar"       3 hours ago   Up 3 hours             k8s_dashboard-metrics-scraper_dashboard-metrics-scraper-7b59f7d4df-xqb96_kubernetes-dashboard_7435c1e5-359b-4350-8fa7-44facdc33e63_8
    fc7a1d65ad43   registry.aliyuncs.com/google_containers/pause:3.2   "/pause"                 3 hours ago   Up 3 hours             k8s_POD_dashboard-metrics-scraper-7b59f7d4df-xqb96_kubernetes-dashboard_7435c1e5-359b-4350-8fa7-44facdc33e63_8
    [root@k8s-master ~]#
    [root@k8s-master ~]# docker ps | grep etcd
    4f0eadebcb08   0369cf4303ff                                        "etcd --advertise-cl…"   3 hours ago   Up 3 hours             k8s_etcd_etcd-k8s-master_kube-system_729d533a85352265aebac662f2565332_12
    e5d19e95ec85   registry.aliyuncs.com/google_containers/pause:3.2   "/pause"                 3 hours ago   Up 3 hours             k8s_POD_etcd-k8s-master_kube-system_729d533a85352265aebac662f2565332_12
    [root@k8s-master ~]#
    
    [root@k8s-node1 ~]# docker ps | grep web
    a049aebee4bb   nginx                                               "/docker-entrypoint.…"   25 minutes ago   Up 25 minutes             k8s_nginx_web-6f7dfb48c4-gmr8d_default_8ed106ee-b02e-4f26-a7c4-c4f014e4c192_0
    0330feccf8b4   registry.aliyuncs.com/google_containers/pause:3.2   "/pause"                 26 minutes ago   Up 26 minutes             k8s_POD_web-6f7dfb48c4-gmr8d_default_8ed106ee-b02e-4f26-a7c4-c4f014e4c192_0
    [root@k8s-node1 ~]#

**@Pod资源共享举例(网络)**

创建一个包含两个容器的Pod

注意：单独创建Pod可以使用命令例如 **kubectl run** nginx --image=nginx 或 yaml文件，这里举例使用yaml文件。一般不单独创建Pod，而是通过控制器创建。

#清理deployment
    [root@k8s-master ~]# kubectl delete deploy --all
    deployment.apps "web" deleted
    [root@k8s-master ~]#
    
    #编辑Pod的yaml文件
    [root@k8s-master ~]# cat test-netshare.yaml
    apiVersion: v1
    kind: Pod
    metadata:
      labels:
        app: test
      name: test-netshare
      namespace: default
    spec:
      containers:
        - image: busybox
          name: hello
          command: ["/bin/sh","-c",'echo "Hello, Kubernetes!" && sleep 3600']
        - image: nginx
          name: web
    [root@k8s-master ~]#
    
    #创建Pod
    [root@k8s-master ~]# kubectl apply -f test-netshare.yaml
    pod/test-netshare created
    [root@k8s-master ~]#

kubectl get pod -o wide可以看到Pod的IP

kubectl describe pod <pod name>，可以看到test-netshare这个Pod里有两个容器，hello和web

[root@k8s-master ~]# kubectl get pod -o wide
    NAME            READY   STATUS    RESTARTS   AGE     IP              NODE        NOMINATED NODE   READINESS GATES
    test-netshare   2/2     Running   0          2m49s   10.244.36.107   k8s-node1   <none>           <none>
    [root@k8s-master ~]#
    [root@k8s-master ~]# kubectl describe pod test-netshare
    Name:         test-netshare
    ……
    IPs:
      IP:  10.244.36.107
    Containers:
      hello:
        Container ID:  docker://73b6704f226233c1776b1fc4ab66a1e2b6326b108f77e5dc1d57b338b68193d8
        Image:         busybox
    ……
      web:
        Container ID:   docker://459f0e2bf7bc33f66dfc523ec3210229783c86ddb27fde309224b481a3219502
        Image:          nginx
    ……

首先进入容器hello看看，注意ipaddr输出的结果，和上述分配给Pod的IP一致

[root@k8s-master ~]# kubectl exec -it test-netshare -c hello -- sh
    / # ip addr
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
        inet 127.0.0.1/8 scope host lo
           valid_lft forever preferred_lft forever
    2: tunl0@NONE: <NOARP> mtu 1480 qdisc noop qlen 1000
        link/ipip 0.0.0.0 brd 0.0.0.0
    4: eth0@if19: <BROADCAST,MULTICAST,UP,LOWER_UP,M-DOWN> mtu 1440 qdisc noqueue
        link/ether 96:02:7f:2d:9c:f0 brd ff:ff:ff:ff:ff:ff
        inet 10.244.36.107/32 scope global eth0
           valid_lft forever preferred_lft forever
    / #

由于另一个容器web用的nginx镜像，这里wget一把，能够下载到nginx的主页

[root@k8s-master ~]# kubectl exec -it test-netshare -c hello -- sh
    / # curl 127.0.0.1:80
    sh: curl: not found
    / # wget 127.0.0.1:80
    Connecting to 127.0.0.1:80 (127.0.0.1:80)
    saving to 'index.html'
    index.html           100% |*********************************************************************************************************|   612  0:00:00 ETA
    'index.html' saved
    / # cat index.html
    <!DOCTYPE html>
    <html>
    <head>
    <title>Welcome to nginx!</title>
    ……
    </html>
    / #

说明：web容器不支持查看网络的命令

[root@k8s-master ~]# kubectl exec -it test-netshare -c web -- sh
    #
    # ip addr
    sh: 2: ip: not found
    # ifconfig
    sh: 3: ifconfig: not found
    # cd /etc/sysconfig
    sh: 4: cd: can't cd to /etc/sysconfig

**@Pod资源共享举例(存储)**

创建一个包含两个容器的Pod

[root@k8s-master ~]# cat test-volshare.yaml
    apiVersion: v1
    kind: Pod
    metadata:
      labels:
        app: test
      name: test-volshare
      namespace: default
    spec:
      containers:
        - image: busybox
          name: hello
          command: ["/bin/sh","-c",'echo "Hello, Kubernetes!" && sleep 3600']
          volumeMounts:
            - name: log
              mountPath: /var/data
        - image: nginx
          name: web
          volumeMounts:
          - name: log
            mountPath: /usr/share/data
      volumes:
        - name: log
          emptyDir: {}
    [root@k8s-master ~]#
    [root@k8s-master ~]# kubectl apply -f test-volshare.yaml
    pod/test-volshare created
    [root@k8s-master ~]#

先进入容器hello，在其数据卷挂载的路径下创建一个文件

[root@k8s-master ~]# kubectl get pod
    NAME            READY   STATUS    RESTARTS   AGE
    test-volshare   2/2     Running   0          67s
    [root@k8s-master ~]#
    [root@k8s-master ~]# kubectl exec -it test-volshare -c hello -- sh
    / # cd /var/data
    /var/data # ls
    /var/data # touch createbyhello
    /var/data # echo 'hello' >> createbyhello
    /var/data # exit
    [root@k8s-master ~]#

然后进入容器web，在其数据卷挂载的路径下可以看到刚才在hello容器中创建的文件

[root@k8s-master ~]# kubectl exec -it test-volshare -c web -- bash
    root@test-volshare:/# cd /usr/share/data/
    root@test-volshare:/usr/share/data# ls
    createbyhello
    root@test-volshare:/usr/share/data# cat createbyhello
    hello
    root@test-volshare:/usr/share/data#