Kubernetes持久化存储2——探究实验

目录贴：Kubernetes学习系列

一、简介

　　本文在“创建PV，创建PVC挂载PV，创建POD挂载PVC”这个环境的基础上，进行各种删除实验，并记录、分析各资源的状态。

二、实验脚本

　　实验创建了一个PV、一个PVC挂载了PV、一个POD挂载PVC，并编写了两个简单的小脚本来快速创建和删除环境。对应的脚本如下所示：

　　需要注意的是在创建PV时，PV并不会去检查你配置的server是否真的存在；也不会检查server上是否有一个可用的NFS服务；当然更不会检查你设置的storage大小是否真有那么大。个人感觉PV的设置只是一个声明，系统并不会对此做任何检查。PVC的挂载也只是根据配额的大小和访问模式，过滤一下PV，并以最小的代价支持。

[root@k8s-master pv]# cat nfs-pv.yaml 
apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv0001
spec:
  capacity:
    storage: 5Gi 
  accessModes:
    - ReadWriteMany 
  persistentVolumeReclaimPolicy: Recycle
  nfs:
    path: "/data/disk1"
    server: 192.168.20.47 
    readOnly: false
[root@k8s-master pv]# cat pvc.yaml 
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: nfs-pvc
spec:
  accessModes:
  - ReadWriteMany      
  resources:
     requests:
       storage: 1Gi     
[root@k8s-master pv]# cat test-pvc-pod.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: test-nfs-pvc
  labels:
    name: test-nfs-pvc
spec:
  containers:
    - name: test-nfs-pvc
      image: registry:5000/back_demon:1.0  
      ports:
        - name: backdemon
          containerPort: 80
      command:
        - /run.sh
      volumeMounts:
        - name: nfs-vol 
          mountPath: /home/laizy/test/nfs-pvc 
  volumes:
    - name: nfs-vol
      persistentVolumeClaim:
        claimName: nfs-pvc
[root@k8s-master pv]# cat start.sh 
#!/bin/bash
kubectl create -f nfs-pv.yaml
kubectl create -f pvc.yaml
kubectl create -f test-pvc-pod.yaml
[root@k8s-master pv]# cat remove.sh 
#!/bin/bash
kubectl delete pod test-nfs-pvc
kubectl delete persistentvolumeclaim nfs-pvc
kubectl delete persistentvolume pv0001
[root@k8s-master pv]#

三、删除PV实验

　　创建PV，创建PVC挂载PV，创建POD挂载PVC。在删除PV后，PVC状态从Bound变为Lost，Pod中的Volume仍然能用，数据也没有被删除。

[root@k8s-master pv]# ./start.sh 
persistentvolume "pv0001" created
persistentvolumeclaim "nfs-pvc" created
pod "test-nfs-pvc" created
[root@k8s-master pv]# kubectl get persistentvolume
NAME   CAPACITY   ACCESSMODES   RECLAIMPOLICY STATUS   CLAIM        REASON    AGE
pv0001   5Gi        RWX           Recycle       Bound   default/nfs-pvc            15s
[root@k8s-master pv]# kubectl get persistentvolumeclaim
NAME    STATUS    VOLUME   CAPACITY   ACCESSMODES   AGE
nfs-pvc   Bound     pv0001    5Gi        RWX           18s
[root@k8s-master pv]# kubectl get pod test-nfs-pvc 
NAME       READY     STATUS    RESTARTS   AGE
test-nfs-pvc   1/1       Running    0          39s
[root@k8s-master pv]# kubectl delete persistentvolume pv0001 
persistentvolume "pv0001" deleted
[root@k8s-master pv]# kubectl get persistentvolume
No resources found.
[root@k8s-master pv]# kubectl get persistentvolumeclaim 
NAME    STATUS    VOLUME    CAPACITY   ACCESSMODES   AGE
nfs-pvc    Lost      pv0001     0                        1m
[root@k8s-master pv]# kubectl get pod test-nfs-pvc 
NAME           READY     STATUS    RESTARTS   AGE
test-nfs-pvc   1/1       Running   0          1m
[root@k8s-master pv]# kubectl exec -ti test-nfs-pvc /bin/bash
[root@test-nfs-pvc /]# cd /home/laizy/test/nfs-pvc/
[root@test-nfs-pvc nfs-pvc]# ls
2.out
[root@test-nfs-pvc nfs-pvc]# exit
exit
[root@k8s-master pv]#

四、删除PVC实验

4.1清空Pod设置

　　设置PV时，如果设置了回收策略是“回收”的时候，在删除PVC时，系统（Controller-Manager）会启动一个recycler的Pod，用于清理数据卷中的内容。每种数据卷的回收Pod是不同的，都有自己特定的逻辑。本文以NFS为例，给出具体配置及Pod描述如下：

[root@k8s-master ~]# cat /etc/kubernetes/controller-manager   #配置完成后请重启controller-manager
###
# The following values are used to configure the kubernetes controller-manager

# defaults from config and apiserver should be adequate

# Add your own!
KUBE_CONTROLLER_MANAGER_ARGS="--pv-recycler-pod-template-filepath-nfs=/etc/kubernetes/recycler.yaml"
[root@k8s-master ~]# cat /etc/kubernetes/recycler.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pv-recycler-
  namespace: default
spec:
  restartPolicy: Never
  volumes:
  - name: vol
    hostPath:
      path: /any/path/it/will/be/replaced
  containers:
  - name: pv-recycler
    image: "docker.io/busybox"
    imagePullPolicy: IfNotPresent
    command: ["/bin/sh", "-c", "test -e /scrub && rm -rf /scrub/..?* /scrub/.[!.]* /scrub/*  && test -z "$(ls -A /scrub)" || exit 1"]
    volumeMounts:
    - name: vol
      mountPath: /scrub
[root@k8s-master ~]#

　　如果不进行上面的设置的话，默认回收Pod用的image是gcr上的busybox，因为种种原因，在国内是无法下载的（即使你的机器上有gcr的busybox也不可以，还需要设置镜像下载策略为IfNotPresent，否则会一直去gcr查询是否有新版本的镜像，这也会导致imagePullError）。所以必须要在Controller-Manager中进行设置，设置成随便哪个busybox。

4.2删除实验

创建PV，创建PVC挂载PV，创建POD挂载PVC。删除PVC后，PV状态从Bound变为Available，系统（controller-manager）调用持久化存储清理插件（recycler-for-pv0001），将PVC对应的PV清空。Pod中的Volume仍然能用，但volume中的数据被删除了。

[root@k8s-master pv]# ./start.sh 
persistentvolume "pv0001" created
persistentvolumeclaim "nfs-pvc" created
pod "test-nfs-pvc" created
[root@k8s-master pv]# kubectl get persistentvolume
NAME      CAPACITY   ACCESSMODES   RECLAIMPOLICY   STATUS    CLAIM             REASON    AGE
pv0001    5Gi        RWX           Recycle         Bound     default/nfs-pvc             11s
[root@k8s-master pv]# kubectl get persistentvolumeclaim 
NAME      STATUS    VOLUME    CAPACITY   ACCESSMODES   AGE
nfs-pvc   Bound     pv0001    5Gi        RWX           14s
[root@k8s-master pv]# kubectl get pod
NAME          READY     STATUS    RESTARTS   AGE
test-nfs-pvc      1/1       Running   0          19s
[root@k8s-master pv]# kubectl exec -ti test-nfs-pvc /bin/bash
[root@test-nfs-pvc /]# touch /home/laizy/test/nfs-pvc/1.out
[root@test-nfs-pvc /]# cd /home/laizy/test/nfs-pvc/
[root@test-nfs-pvc nfs-pvc]# ls
1.out
[root@test-nfs-pvc nfs-pvc]# exit
exit
[root@k8s-master pv]# kubectl delete persistentvolumeclaim nfs-pvc 
persistentvolumeclaim "nfs-pvc" deleted
[root@k8s-master pv]# kubectl get pod
NAME                 READY     STATUS              RESTARTS   AGE
recycler-for-pv0001       0/1       ContainerCreating    0          1s
test-nfs-pvc             1/1       Running             0          1m
[root@k8s-master pv]# kubectl get persistentvolume
NAME      CAPACITY   ACCESSMODES   RECLAIMPOLICY   STATUS      CLAIM     REASON    AGE
pv0001    5Gi        RWX           Recycle         Available                       1m
[root@k8s-master pv]# kubectl get persistentvolumeclaim 
No resources found.
[root@k8s-master pv]# kubectl get pod test-nfs-pvc 
NAME           READY     STATUS    RESTARTS   AGE
test-nfs-pvc   1/1       Running   0          2m
[root@k8s-master pv]# kubectl exec -ti test-nfs-pvc /bin/bash
[root@test-nfs-pvc /]# ls /home/laizy/test/nfs-pvc/
[root@test-nfs-pvc /]

五、删除Pod实验

　　创建PV，创建PVC挂载PV，创建POD挂载PVC。删除Pod后，PV、PVC状态没变，Pod中的Volume对应的NFS数据没有被删除。

[root@k8s-master pv]# ./start.sh 
persistentvolume "pv0001" created
persistentvolumeclaim "nfs-pvc" created
pod "test-nfs-pvc" created
[root@k8s-master pv]# kubectl get persistentvolume
NAME  CAPACITY ACCESSMODES   RECLAIMPOLICY   STATUS    CLAIM   REASON    AGE
pv0001 5Gi      RWX           Recycle         Bound     default/nfs-pvc       11s
[root@k8s-master pv]# kubectl get persistentvolumeclaim
NAME      STATUS    VOLUME    CAPACITY   ACCESSMODES   AGE
nfs-pvc   Bound     pv0001    5Gi        RWX           27s
[root@k8s-master pv]# kubectl get pod
NAME         READY     STATUS    RESTARTS   AGE
test-nfs-pvc     1/1       Running   0          36s
[root@k8s-master pv]# kubectl exec -ti test-nfs-pvc /bin/bash
[root@test-nfs-pvc /]# cat /home/laizy/test/nfs-pvc/1.out 
123456
[root@test-nfs-pvc /]# exit
exit
[root@k8s-master pv]# kubectl delete pod test-nfs-pvc 
pod "test-nfs-pvc" deleted
[root@k8s-master pv]# kubectl get persistentvolume
NAME CAPACITY   ACCESSMODES   RECLAIMPOLICY   STATUS    CLAIM  REASON    AGE
pv0001    5Gi   RWX           Recycle         Bound     default/nfs-pvc      8m
[root@k8s-master pv]# kubectl get persistentvolumeclaim
NAME      STATUS    VOLUME    CAPACITY   ACCESSMODES   AGE
nfs-pvc     Bound     pv0001     5Gi        RWX           8m
[root@k8s-master pv]# ssh 192.168.20.47  #登录到远程NFS服务器
root@192.168.20.47's password: 
Last failed login: Mon Mar 27 14:37:19 CST 2017 from :0 on :0
There was 1 failed login attempt since the last successful login.
Last login: Mon Mar 20 10:49:18 2017
[root@localhost ~]# cd /data/disk1/1.out
123456