k8s持久化存储PV、PVC、StorageClass

k8s持久化存储

1. 以前数据持久化方式

  通过volumes 数据卷挂载

1. web3.yaml 内容如下:

apiVersion: apps/v1
kind: Deployment
metadata:
  creationTimestamp: null
  labels:
    app: web3
  name: web3
spec:
  replicas: 1
  selector:
    matchLabels:
      app: web3
  strategy: {}
  template:
    metadata:
      creationTimestamp: null
      labels:
        app: web3
    spec:
      containers:
      - image: nginx
        name: nginx
        resources: {}
        volumeMounts:
        - name: varlog
          mountPath: /tmp/log
      volumes:
      - name: varlog
        hostPath:
          path: /tmp/log/web3log
status: {}

2. 创建资源后查看

[root@k8smaster1 volumestest]# kubectl get pods | grep web3
web3-6c6557674d-xt7kr                           1/1     Running   0          6m38s
[root@k8smaster1 volumestest]# kubectl describe pods web3-6c6557674d-xt7kr

相关挂载信息如下:

 3. 到容器内部创建一个文件

[root@k8smaster1 volumestest]# kubectl exec -it web3-6c6557674d-xt7kr
error: you must specify at least one command for the container
[root@k8smaster1 volumestest]# kubectl exec -it web3-6c6557674d-xt7kr -- bash
root@web3-6c6557674d-xt7kr:/# echo "123" > /tmp/log/test.txt
root@web3-6c6557674d-xt7kr:/# exit
exit

4. 到pod 调度的节点查看宿主机目录是否挂载成功

(1) master 节点查看pod 调度节点

[root@k8smaster1 volumestest]# kubectl get pods -o wide | grep web3
web3-6c6557674d-xt7kr                           1/1     Running   0          11m     10.244.2.108     k8snode2     <none>           <none>

(2) 到k8snode2 节点查看

[root@k8snode2 web3log]# ll
total 4
-rw-r--r-- 1 root root 4 Jan 21 05:49 test.txt
[root@k8snode2 web3log]# cat test.txt 
123

5. 测试k8snode2 节点宕机,pod 自动调度到k8snode1 节点再次查看

[root@k8smaster1 volumestest]# kubectl get pods -o wide | grep web3
web3-6c6557674d-6wlh4                           1/1     Running       0          4m22s   10.244.1.110     k8snode1     <none>           <none>
web3-6c6557674d-xt7kr                           1/1     Terminating   0          22m     10.244.2.108     k8snode2     <none>           <none>
[root@k8smaster1 volumestest]# kubectl exec -it web3-6c6557674d-6wlh4 -- bash
root@web3-6c6557674d-6wlh4:/# ls /tmp/log/
root@web3-6c6557674d-6wlh4:/# 

   发现自动调度到k8snode1 节点,进入容器之后发现之前新建的文件丢失。

6. 从k8snode1 宿主机查看发现也没有文件

[root@k8snode1 web3log]# pwd
/tmp/log/web3log
[root@k8snode1 web3log]# ls
[root@k8snode1 web3log]# 

  造成的现象就是pod 所在的节点宕机后,volume 数据卷挂载的文件也丢失,因此需要一种解决方案。

1. nfs 持久化存储

  网络文件系统,是一种共享文件系统,实际上相当于客户端将文件上传到服务器,实现共享。

1. 下载nfs

1. 找一台服务器安装nfs

(1) 安装nfs以及查看nfs 服务状态

yum install -y nfs-utils

(2) 设置挂载路径, 注意需要将挂载路径创建出来

[root@k8smaster2 logs]# cat /etc/exports
/data/nfs *(rw,no_root_squash)

解释: rw 代表读写访问, no_root_squash 代表root 用户具有根目录的完全管理访问权限

2. k8s 集群node 节点安装nfs-utils

yum install -y nfs-utils

3. nfs 服务器启动nfs 服务且查看服务状态

[root@k8smaster2 nfs]# systemctl start nfs    # 启动nfs
[root@k8smaster2 nfs]# systemctl status nfs    # 查看状态
● nfs-server.service - NFS server and services
   Loaded: loaded (/usr/lib/systemd/system/nfs-server.service; disabled; vendor preset: disabled)
   Active: active (exited) since Fri 2022-01-21 19:55:38 EST; 5min ago
  Process: 51947 ExecStopPost=/usr/sbin/exportfs -f (code=exited, status=0/SUCCESS)
  Process: 51943 ExecStopPost=/usr/sbin/exportfs -au (code=exited, status=0/SUCCESS)
  Process: 51941 ExecStop=/usr/sbin/rpc.nfsd 0 (code=exited, status=0/SUCCESS)
  Process: 51977 ExecStartPost=/bin/sh -c if systemctl -q is-active gssproxy; then systemctl reload gssproxy ; fi (code=exited, status=0/SUCCESS)
  Process: 51960 ExecStart=/usr/sbin/rpc.nfsd $RPCNFSDARGS (code=exited, status=0/SUCCESS)
  Process: 51958 ExecStartPre=/usr/sbin/exportfs -r (code=exited, status=0/SUCCESS)
 Main PID: 51960 (code=exited, status=0/SUCCESS)
    Tasks: 0
   Memory: 0B
   CGroup: /system.slice/nfs-server.service

Jan 21 19:55:38 k8smaster2 systemd[1]: Starting NFS server and services...
Jan 21 19:55:38 k8smaster2 systemd[1]: Started NFS server and services.
[root@k8smaster2 nfs]# showmount -e localhost    # 查看挂载的nfs 信息
Export list for localhost:
/data/nfs *

也可以查看nfs 的进程信息

[root@k8smaster2 nfs]# ps -ef | grep nfs
root      51962      2  0 19:55 ?        00:00:00 [nfsd4_callbacks]
root      51968      2  0 19:55 ?        00:00:00 [nfsd]
root      51969      2  0 19:55 ?        00:00:00 [nfsd]
root      51970      2  0 19:55 ?        00:00:00 [nfsd]
root      51971      2  0 19:55 ?        00:00:00 [nfsd]
root      51972      2  0 19:55 ?        00:00:00 [nfsd]
root      51973      2  0 19:55 ?        00:00:00 [nfsd]
root      51974      2  0 19:55 ?        00:00:00 [nfsd]
root      51975      2  0 19:55 ?        00:00:00 [nfsd]
root      54774  45013  0 20:02 pts/2    00:00:00 grep --color=auto nfs

2. 客户端安装

 1. 在所有k8snode 节点安装客户端,并且查看远程nfs 信息

yum install -y nfs-utils

2. 查看远程信息

[root@k8snode1 ~]# showmount -e 192.168.13.106
Export list for 192.168.13.106:
/data/nfs *

3. 本地测试nfs

(1) 创建挂载并进行测试

[root@k8snode1 ~]# mkdir /share
[root@k8snode1 ~]# mount 192.168.13.106:/data/nfs /share
[root@k8snode1 ~]# df -h | grep 13.106
192.168.13.106:/data/nfs   17G   12G  5.4G  69% /share

(2) node 节点创建文件

[root@k8snode1 ~]# echo "hello from 104" >> /share/104.txt
[root@k8snode1 ~]# cat /share/104.txt 
hello from 104

(3) nfs 服务器查看

[root@k8smaster2 nfs]# cat 104.txt 
hello from 104

(4) 客户端取消挂载

[root@k8snode1 ~]# umount /share
[root@k8snode1 ~]# df -h | grep 13.106

取消挂载之后,nfs 服务器上的文件仍然存在。

3. k8s 集群使用nfs

1. 编写nfs-nginx.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-dep1
spec:
  replicas: 1
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        volumeMounts:
        - name: wwwroot
          mountPath: /usr/share/nginx/html
        ports:
        - containerPort: 80
      volumes:
        - name: wwwroot
          nfs:
            server: 192.168.13.106
            path: /data/nfs

2. 创建资源

[root@k8smaster1 nfs]# kubectl apply -f nfs-nginx.yaml 
deployment.apps/nginx-dep1 created

然后查看pod describe 信息

3. 我们进入容器然后创建一个文件导/usr/share/nginx/html

root@nginx-dep1-6d7f9c85dc-lqfbf:/# cat /usr/share/nginx/html/index.html 
hello

4. 然后到nfs 服务器查看

[root@k8smaster2 nfs]# pwd
/data/nfs
[root@k8smaster2 nfs]# ls
104.txt  index.html
[root@k8smaster2 nfs]# cat index.html 
hello

4. pv 和 pvc

上面使用nfs 有一个问题,就是每个需要持久化的都需要知道远程nfs 服务器的地址以及相关权限,可能不太安全。下面研究pv和pvc 使用。

pv pvc 对应PersistentVolume和PersistentVolumeClaim。  pv 类似于一个声明nfs 地址等信息,抽象成配置文件; pvc 通过引用pv 中声明的信息,然后即可实现nfs 持久化存储。

pv 有好多实现方式,实际上是对nfs进行一层包装,因为我们已经安装了nfs, 所以基于nfs 实现。

参考: https://kubernetes.io/docs/concepts/storage/persistent-volumes/

1.  创建pv

1. 创建 pv.yaml

apiVersion: v1
kind: PersistentVolume
metadata:
  name: my-pv
spec:
  capacity:
    storage: 5Gi
  accessModes:
    - ReadWriteMany
  nfs:
    path: /data/nfs
    server: 192.168.13.106

2. 创建并查看

[root@k8smaster1 nfs]# kubectl apply -f pv.yaml 
persistentvolume/my-pv created
[root@k8smaster1 nfs]# kubectl get pv -o wide
NAME    CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM   STORAGECLASS   REASON   AGE    VOLUMEMODE
my-pv   5Gi        RWX            Retain           Available                                   2m4s   Filesystem

补充: 关于PV的一些核心概念

1. test-pv.yml

apiVersion: v1
kind: PersistentVolume
metadata:
  name:  pv2
spec:
  capacity: 
    storage: 1Gi
  accessModes:
  - ReadWriteOnce
  persistentVolumeReclaimPolicy: Recycle
  nfs:
    path: /data/nfs
    server: 192.168.13.106

2. 执行创建并且查看

[root@k8smaster1 storageclass]# kubectl apply -f test-pv.yml 
persistentvolume/pv2 created
[root@k8smaster1 storageclass]# kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM              STORAGECLASS         REASON   AGE
pv2                                        1Gi        RWO            Recycle          Available                                                    5s
pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            Delete           Bound       default/test-pvc   course-nfs-storage            27m
[root@k8smaster1 storageclass]# kubectl describe pv pv2
Name:            pv2
Labels:          <none>
Annotations:     Finalizers:  [kubernetes.io/pv-protection]
StorageClass:    
Status:          Available
Claim:           
Reclaim Policy:  Recycle
Access Modes:    RWO
VolumeMode:      Filesystem
Capacity:        1Gi
Node Affinity:   <none>
Message:         
Source:
    Type:      NFS (an NFS mount that lasts the lifetime of a pod)
    Server:    192.168.13.106
    Path:      /data/nfs
    ReadOnly:  false
Events:        <none>

3. 核心概念

(1) Capacity(存储能力)

  一般来说,一个 PV 对象都要指定一个存储能力,通过 PV 的 capacity属性来设置的,目前只支持存储空间的设置,就是我们这里的 storage=1Gi,不过未来可能会加入 IOPS、吞吐量等指标的配置。

(2) AccessModes(访问模式)

AccessModes 是用来对 PV 进行访问模式的设置,用于描述用户应用对存储资源的访问权限,访问权限包括下面几种方式:

ReadWriteOnce(RWO):读写权限,但是只能被单个节点挂载

ReadOnlyMany(ROX):只读权限,可以被多个节点挂载

ReadWriteMany(RWX):读写权限,可以被多个节点挂载

(3) persistentVolumeReclaimPolicy(回收策略)

我这里指定的 PV 的回收策略为 Recycle,目前 PV 支持的策略有三种:

Retain(保留)- 保留数据,需要管理员手工清理数据

Recycle(回收)- 清除 PV 中的数据,效果相当于执行 rm -rf /thevoluem/*

Delete(删除)- 与 PV 相连的后端存储完成 volume 的删除操作,当然这常见于云服务商的存储服务,比如 ASW EBS。

(4) 状态:

一个 PV 的生命周期中,可能会处于4种不同的阶段:

Available(可用):表示可用状态,还未被任何 PVC 绑定

Bound(已绑定):表示 PVC 已经被 PVC 绑定

Released(已释放):PVC 被删除,但是资源还未被集群重新声明

Failed(失败): 表示该 PV 的自动回收失败

2. 创建pvc 使用上面的pv

1. 创建pvc.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-dep1
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        volumeMounts:
        - name: wwwroot
          mountPath: /usr/share/nginx/html
        ports:
        - containerPort: 80
      volumes:
      - name: wwwroot
        persistentVolumeClaim:
          claimName: my-pvc

---

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: my-pvc
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 5Gi

2. 创建并查看

[root@k8smaster1 nfs]# kubectl apply -f pvc.yaml 
deployment.apps/nginx-dep1 created
persistentvolumeclaim/my-pvc created
[root@k8smaster1 nfs]# kubectl get pvc -o wide
NAME     STATUS   VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE   VOLUMEMODE
my-pvc   Bound    my-pv    5Gi        RWX                           60s   Filesystem
[root@k8smaster1 nfs]# kubectl get pods -o wide
NAME                                            READY   STATUS    RESTARTS   AGE     IP               NODE         NOMINATED NODE   READINESS GATES
nginx-dep1-58b7bf955f-4jhbq                     1/1     Running   0          75s     10.244.2.112     k8snode2     <none>           <none>
nginx-dep1-58b7bf955f-m69dm                     1/1     Running   0          75s     10.244.2.110     k8snode2     <none>           <none>
nginx-dep1-58b7bf955f-qh6pg                     1/1     Running   0          75s     10.244.2.111     k8snode2     <none>           <none>
nginx-f89759699-vkf7d                           1/1     Running   3          4d16h   10.244.1.106     k8snode1     <none>           <none>
tomcat-58767d5b5-f5qwj                          1/1     Running   2          4d15h   10.244.1.103     k8snode1     <none>           <none>
weave-scope-agent-ui-kbq7b                      1/1     Running   2          45h     192.168.13.105   k8snode2     <none>           <none>
weave-scope-agent-ui-tg5q4                      1/1     Running   2          45h     192.168.13.103   k8smaster1   <none>           <none>
weave-scope-agent-ui-xwh2b                      1/1     Running   2          45h     192.168.13.104   k8snode1     <none>           <none>
weave-scope-cluster-agent-ui-7498b8d4f4-zdlk7   1/1     Running   2          45h     10.244.1.104     k8snode1     <none>           <none>
weave-scope-frontend-ui-649c7dcd5d-7gb9s        1/1     Running   2          45h     10.244.1.107     k8snode1     <none>           <none>
web3-6c6557674d-6wlh4                           1/1     Running   0          14h     10.244.1.110     k8snode1     <none>           <none>
[root@k8smaster1 nfs]# 

3. 随便进入一个pod的第一个容器,然后创建文件

[root@k8smaster1 nfs]# kubectl exec -it nginx-dep1-58b7bf955f-4jhbq -- bash
root@nginx-dep1-58b7bf955f-4jhbq:/# echo "111222" >> /usr/share/nginx/html/1.txt
root@nginx-dep1-58b7bf955f-4jhbq:/# exit
exit

4. 到nfs 服务器查看与其他容器查看

(1) nfs 服务器查看

[root@k8smaster2 nfs]# ls
104.txt  1.txt  index.html
[root@k8smaster2 nfs]# cat 1.txt 
111222

(2) 进入其他pod 的第一个容器查看

[root@k8smaster1 nfs]# kubectl exec -it nginx-dep1-58b7bf955f-qh6pg -- bash
root@nginx-dep1-58b7bf955f-qh6pg:/# ls /usr/share/nginx/html/
1.txt  104.txt  index.html

  至此简单实现了基于nfs 和 pv、pvc 的持久化存储。

5. storageclass

  PV 可以理解为静态的,就是要使用的一个 PVC 的话就必须手动去创建一个 PV,这种方式在很大程度上并不能满足我们的需求,比如我们有一个应用需要对存储的并发度要求比较高,而另外一个应用对读写速度又要求比较高,特别是对于 StatefulSet 类型的应用简单的来使用静态的 PV 就很不合适了,这种情况下我们就需要用到动态 PV,也就 StorageClass。

1. 创建storageclass

  要使用 StorageClass,我们就得安装对应的自动配置程序,比如我们这里存储后端使用的是 nfs,那么我们就需要使用到一个 nfs-client 的自动配置程序,我们也叫它 Provisioner,这个程序使用我们已经配置好的 nfs 服务器,来自动创建持久卷,也就是自动帮我们创建 PV。

  自动创建的 PV 以${namespace}-${pvcName}-${pvName}这样的命名格式创建在 NFS 服务器上的共享数据目录中,而当这个 PV 被回收后会以archieved-${namespace}-${pvcName}-${pvName}这样的命名格式存在 NFS 服务器上。

  当然在部署nfs-client之前,我们需要先成功安装上 nfs 服务器,服务地址是192.168.13.106,共享数据目录是/data/nfs/,然后接下来我们部署 nfs-client 即可,我们也可以直接参考 nfs-client 的文档:https://github.com/kubernetes-incubator/external-storage/tree/master/nfs-client ,进行安装即可。

第一步: 配置 Deployment,将里面的对应的参数替换成我们自己的 nfs 配置(nfs-client.yml)

kind: Deployment
apiVersion: apps/v1
metadata:
  name: nfs-client-provisioner
spec:
  replicas: 1
  strategy:
    type: Recreate
  selector:
    matchLabels:
      app: nfs-client-provisioner
  template:
    metadata:
      labels:
        app: nfs-client-provisioner
    spec:
      serviceAccountName: nfs-client-provisioner
      containers:
        - name: nfs-client-provisioner
          image: quay.io/external_storage/nfs-client-provisioner:latest
          volumeMounts:
            - name: nfs-client-root
              mountPath: /persistentvolumes
          env:
            - name: PROVISIONER_NAME
              value: fuseim.pri/ifs
            - name: NFS_SERVER
              value: 192.168.13.106
            - name: NFS_PATH
              value: /data/nfs
      volumes:
        - name: nfs-client-root
          nfs:
            server: 192.168.13.106
            path: /data/nfs
View Code

第二步:使用一个名为 nfs-client-provisioner 的serviceAccount,也需要创建一个 sa,然后绑定上对应的权限:(nfs-client-sa.yml)

apiVersion: v1
kind: ServiceAccount
metadata:
  name: nfs-client-provisioner

---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: nfs-client-provisioner-runner
rules:
  - apiGroups: [""]
    resources: ["persistentvolumes"]
    verbs: ["get", "list", "watch", "create", "delete"]
  - apiGroups: [""]
    resources: ["persistentvolumeclaims"]
    verbs: ["get", "list", "watch", "update"]
  - apiGroups: ["storage.k8s.io"]
    resources: ["storageclasses"]
    verbs: ["get", "list", "watch"]
  - apiGroups: [""]
    resources: ["events"]
    verbs: ["list", "watch", "create", "update", "patch"]
  - apiGroups: [""]
    resources: ["endpoints"]
    verbs: ["create", "delete", "get", "list", "watch", "patch", "update"]

---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: run-nfs-client-provisioner
subjects:
  - kind: ServiceAccount
    name: nfs-client-provisioner
    namespace: default
roleRef:
  kind: ClusterRole
  name: nfs-client-provisioner-runner
  apiGroup: rbac.authorization.k8s.io
View Code

  我们这里新建的一个名为 nfs-client-provisioner 的ServiceAccount,然后绑定了一个名为 nfs-client-provisioner-runner 的ClusterRole,而该ClusterRole声明了一些权限,其中就包括对persistentvolumes的增、删、改、查等权限,所以我们可以利用该ServiceAccount来自动创建 PV。

第三步: nfs-client 的 Deployment 声明完成后,我们就可以来创建一个StorageClass对象了:(nfs-client-class.yml)

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: course-nfs-storage
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
View Code

  声明了一个名为 course-nfs-storage 的StorageClass对象,注意下面的provisioner对应的值一定要和上面的Deployment下面的 PROVISIONER_NAME 这个环境变量的值一样。

接下来使用kubectl apply -f XXX.yml 创建上面资源并且查看相关资源:

[root@k8smaster1 storageclass]# kubectl get pods,deployments -o wide | grep nfs 
pod/nfs-client-provisioner-6888b56547-7ts79         1/1     Running   0          101m   10.244.2.118     k8snode2     <none>           <none>
deployment.apps/nfs-client-provisioner         1/1     1            1           3h26m   nfs-client-provisioner      quay.io/external_storage/nfs-client-provisioner:latest   app=nfs-client-provisioner
[root@k8smaster1 storageclass]# kubectl get storageclass -o wide
NAME                 PROVISIONER      RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
course-nfs-storage   fuseim.pri/ifs   Delete          Immediate           false                  44m

也可以创建的时候设置为默认的storageclass

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: course-nfs-storage
  annotations:
        storageclass.kubernetes.io/is-default-class: "true"
provisioner: fuseim.pri/ifs

 

2. 新建

1. 首先创建一个 PVC 对象, test-pvc.yml

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: test-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 1Mi

执行创建:

[root@k8smaster1 storageclass]# kubectl apply -f test-pvc.yml 
persistentvolumeclaim/test-pvc created
[root@k8smaster1 storageclass]# kubectl get pvc -o wide
NAME       STATUS    VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE   VOLUMEMODE
test-pvc   Pending                                                     2s    Filesystem

  声明了一个 PVC 对象,采用 ReadWriteMany 的访问模式,请求 1Mi 的空间,但是我们可以看到上面的 PVC 文件我们没有标识出任何和 StorageClass 相关联的信息,那么如果我们现在直接创建这个 PVC 对象不会自动绑定上合适的 PV 对象,我们这里有两种方法可以来利用上面我们创建的 StorageClass 对象来自动帮我们创建一个合适的 PV

方法1:我们可以设置这个 course-nfs-storage 的 StorageClass 为 Kubernetes 的默认存储后端

kubectl patch storageclass course-nfs-storage -p '{"metadata": {"annotations":{"storageclass.kubernetes.io/is-default-class":"true"}}}'

查看默认的storageclass以及取消默认:

[root@k8smaster1 storageclass]# kubectl get storageclass
NAME                           PROVISIONER      RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
course-nfs-storage (default)   fuseim.pri/ifs   Delete          Immediate           false                  122m
[root@k8smaster1 storageclass]# kubectl patch storageclass course-nfs-storage -p '{"metadata": {"annotations":{"storageclass.kubernetes.io/is-default-class":"false"}}}'
storageclass.storage.k8s.io/course-nfs-storage patched
[root@k8smaster1 storageclass]# kubectl get storageclass
NAME                 PROVISIONER      RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
course-nfs-storage   fuseim.pri/ifs   Delete          Immediate           false                  123m

方法二:在这个 PVC 对象中添加一个声明 StorageClass 对象的标识,这里我们可以利用一个 annotations 属性来标识,如下: (推荐这种)

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: test-pvc
  annotations:
    volume.beta.kubernetes.io/storage-class: "course-nfs-storage"
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 1Mi

创建并且查看:

[root@k8smaster1 storageclass]# kubectl apply -f test-pvc.yml 
persistentvolumeclaim/test-pvc created
[root@k8smaster1 storageclass]# kubectl get pvc
NAME       STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS         AGE
test-pvc   Bound    pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            course-nfs-storage   5s
[root@k8smaster1 storageclass]# kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM              STORAGECLASS         REASON   AGE
pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            Delete           Bound    default/test-pvc   course-nfs-storage            53s
[root@k8smaster1 storageclass]# kubectl describe pv pvc-97bce597-0788-49a1-be6d-5a938363797b
Name:            pvc-97bce597-0788-49a1-be6d-5a938363797b
Labels:          <none>
Annotations:     pv.kubernetes.io/provisioned-by: fuseim.pri/ifs
Finalizers:      [kubernetes.io/pv-protection]
StorageClass:    course-nfs-storage
Status:          Bound
Claim:           default/test-pvc
Reclaim Policy:  Delete
Access Modes:    RWX
VolumeMode:      Filesystem
Capacity:        1Mi
Node Affinity:   <none>
Message:         
Source:
    Type:      NFS (an NFS mount that lasts the lifetime of a pod)
    Server:    192.168.13.106
    Path:      /data/nfs/default-test-pvc-pvc-97bce597-0788-49a1-be6d-5a938363797b
    ReadOnly:  false
Events:        <none>
[root@k8smaster1 storageclass]# kubectl describe pvc test-pvc
Name:          test-pvc
Namespace:     default
StorageClass:  course-nfs-storage
Status:        Bound
Volume:        pvc-97bce597-0788-49a1-be6d-5a938363797b
Labels:        <none>
Annotations:   pv.kubernetes.io/bind-completed: yes
               pv.kubernetes.io/bound-by-controller: yes
               volume.beta.kubernetes.io/storage-class: course-nfs-storage
               volume.beta.kubernetes.io/storage-provisioner: fuseim.pri/ifs
Finalizers:    [kubernetes.io/pvc-protection]
Capacity:      1Mi
Access Modes:  RWX
VolumeMode:    Filesystem
Mounted By:    <none>
Events:
  Type    Reason                 Age                    From                                                                                         Message
  ----    ------                 ----                   ----                                                                                         -------
  Normal  ExternalProvisioning   5m18s (x2 over 5m18s)  persistentvolume-controller                                                                  waiting for a volume to be created, either by external provisioner "fuseim.pri/ifs" or manually created by system administrator
  Normal  Provisioning           5m18s                  fuseim.pri/ifs_nfs-client-provisioner-6888b56547-7ts79_6aa1d177-8966-11ec-b368-9e5ccaa198de  External provisioner is provisioning volume for claim "default/test-pvc"
  Normal  ProvisioningSucceeded  5m18s                  fuseim.pri/ifs_nfs-client-provisioner-6888b56547-7ts79_6aa1d177-8966-11ec-b368-9e5ccaa198de  Successfully provisioned volume pvc-97bce597-0788-49a1-be6d-5a938363797b

  可以看到: 一个名为 test-pvc 的 PVC 对象创建成功了,状态已经是 Bound 了,也产生了一个对应的 VOLUME 对象,最重要的一栏是 STORAGECLASS,现在也有值了,就是我们刚刚创建的 StorageClass 对象 course-nfs-storage。 并且也自动创建了一个pv 对象,访问模式是 RWX,回收策略是 Delete,这个是通过StorageClass 对象自动创建的。

 3. 测试

1. 新建test-pvc-pod.yml

kind: Pod
apiVersion: v1
metadata:
  name: test-pod
spec:
  containers:
  - name: test-pod
    image: busybox
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    args:
    - "-c"
    - "touch /mnt/SUCCESS && exit 0 || exit 1"
    volumeMounts:
    - name: nfs-pvc
      mountPath: "/mnt"
  restartPolicy: "Never"
  volumes:
  - name: nfs-pvc
    persistentVolumeClaim:
      claimName: test-pvc
View Code

  上面这个 Pod 非常简单,就是用一个 busybox 容器(这个容器集成了常见的linux 命令),在 /mnt 目录下面新建一个 SUCCESS 的文件,然后把 /mnt 目录挂载到上面我们新建的 test-pvc 这个资源对象上面了,要验证很简单,只需要去查看下我们 nfs 服务器上面的共享数据目录下面是否有 SUCCESS 这个文件即可

2. 创建后查看

[root@k8smaster1 storageclass]# kubectl apply -f test-pvc-pod.yml 
pod/test-pod created
[root@k8smaster1 storageclass]# kubectl get pods -o wide | grep test-
test-pod                                        0/1     Completed   0          3m48s   10.244.2.119     k8snode2     <none>           <none>

3. 到nfs 服务器节点查看

[root@k8smaster2 default-test-pvc-pvc-97bce597-0788-49a1-be6d-5a938363797b]# pwd
/data/nfs/default-test-pvc-pvc-97bce597-0788-49a1-be6d-5a938363797b
[root@k8smaster2 default-test-pvc-pvc-97bce597-0788-49a1-be6d-5a938363797b]# ll
total 0
-rw-r--r-- 1 root root 0 Feb  9 02:36 SUCCESS

  可以看到nfs 服务器挂载的目录下面有名字很长的文件夹,这个文件夹的命名满足规则:${namespace}-${pvcName}-${pvName}

4. 常用方法

使用 StorageClass 更多的是 StatefulSet 类型的服务,StatefulSet 类型的服务我们也可以通过一个 volumeClaimTemplates 属性来直接使用 StorageClass

1.  test-statefulset-nfs.yml

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: nfs-web
spec:
  serviceName: "nginx"
  replicas: 3
  selector:
    matchLabels:
      app: nfs-web
  template:
    metadata:
      labels:
        app: nfs-web
    spec:
      terminationGracePeriodSeconds: 10
      containers:
      - name: nginx
        image: nginx:1.7.9
        ports:
        - containerPort: 80
          name: web
        volumeMounts:
        - name: www
          mountPath: /usr/share/nginx/html
  volumeClaimTemplates:
  - metadata:
      name: www
      annotations:
        volume.beta.kubernetes.io/storage-class: course-nfs-storage
    spec:
      accessModes: [ "ReadWriteOnce" ]
      resources:
        requests:
          storage: 1Gi

  实际上 volumeClaimTemplates 下面就是一个 PVC 对象的模板,就类似于我们这里 StatefulSet 下面的 template,实际上就是一个 Pod 的模板,我们不单独创建成 PVC 对象,而用这种模板就可以动态的去创建了对象了,这种方式在 StatefulSet 类型的服务下面使用得非常多。

2. 创建并查看

[root@k8smaster1 storageclass]# kubectl apply -f test-statefulset-nfs.yml 
pstatefulset.apps/nfs-web created
[root@k8smaster1 storageclass]# kubectl get pods | grep nfs-web
nfs-web-0                                       1/1     Running     0          2m42s
nfs-web-1                                       1/1     Running     0          115s
nfs-web-2                                       1/1     Running     0          109s
[root@k8smaster1 storageclass]# kubectl get pvc
NAME            STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS         AGE
test-pvc        Bound    pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            course-nfs-storage   50m
www-nfs-web-0   Bound    pvc-c234c21b-c3c4-4ffb-a14b-aa47cad7183e   1Gi        RWO            course-nfs-storage   2m57s
www-nfs-web-1   Bound    pvc-7fdeb85f-481e-48c1-9734-284cce8014fb   1Gi        RWO            course-nfs-storage   2m10s
www-nfs-web-2   Bound    pvc-7810f38b-2779-49e3-84f2-4b56e16df419   1Gi        RWO            course-nfs-storage   2m4s
[root@k8smaster1 storageclass]# kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                   STORAGECLASS         REASON   AGE
pvc-7810f38b-2779-49e3-84f2-4b56e16df419   1Gi        RWO            Delete           Bound    default/www-nfs-web-2   course-nfs-storage            2m16s
pvc-7fdeb85f-481e-48c1-9734-284cce8014fb   1Gi        RWO            Delete           Bound    default/www-nfs-web-1   course-nfs-storage            2m22s
pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            Delete           Bound    default/test-pvc        course-nfs-storage            50m
pvc-c234c21b-c3c4-4ffb-a14b-aa47cad7183e   1Gi        RWO            Delete           Bound    default/www-nfs-web-0   course-nfs-storage            3m9s

3. 到nfs 服务器查看共享目录如下

[root@k8smaster2 nfs]# pwd
/data/nfs
[root@k8smaster2 nfs]# ll
total 4
drwxrwxrwx 2 root root 21 Feb  9 02:36 default-test-pvc-pvc-97bce597-0788-49a1-be6d-5a938363797b
drwxrwxrwx 2 root root  6 Feb  9 02:49 default-www-nfs-web-0-pvc-c234c21b-c3c4-4ffb-a14b-aa47cad7183e
drwxrwxrwx 2 root root  6 Feb  9 02:50 default-www-nfs-web-1-pvc-7fdeb85f-481e-48c1-9734-284cce8014fb
drwxrwxrwx 2 root root  6 Feb  9 02:50 default-www-nfs-web-2-pvc-7810f38b-2779-49e3-84f2-4b56e16df419
-rw-r--r-- 1 root root  4 Feb  8 21:22 test.txt
[root@k8smaster2 nfs]# 

 补充: StorageClass 相当于一个创建 PV 的模板,用户通过 PVC 申请存储卷,StorageClass 通过模板自动创建 PV,然后和 PVC 进行绑定。

  在有storageclass 环境的k8s 中,可以通过如下方式创建storageclass 以及 pvc、pv

1. yml 内容如下

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: sc2
provisioner: fuseim.pri/ifs
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: test-pvc-sc
  annotations:
    volume.beta.kubernetes.io/storage-class: "sc2"
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 1Mi

2. 执行创建查看资源如下

[root@k8smaster1 storageclass]# kubectl get sc
NAME                 PROVISIONER      RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
course-nfs-storage   fuseim.pri/ifs   Delete          Immediate           false                  169m
sc2                  fuseim.pri/ifs   Delete          Immediate           false                  9m28s
[root@k8smaster1 storageclass]# kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                   STORAGECLASS         REASON   AGE
pvc-1d0c10d4-c7f7-433f-8143-78b11fd8fe58   1Mi        RWX            Delete           Bound    default/test-pvc-sc     sc2                           9m40s
pvc-7810f38b-2779-49e3-84f2-4b56e16df419   1Gi        RWO            Delete           Bound    default/www-nfs-web-2   course-nfs-storage            65m
pvc-7fdeb85f-481e-48c1-9734-284cce8014fb   1Gi        RWO            Delete           Bound    default/www-nfs-web-1   course-nfs-storage            65m
pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            Delete           Bound    default/test-pvc        course-nfs-storage            113m
pvc-c234c21b-c3c4-4ffb-a14b-aa47cad7183e   1Gi        RWO            Delete           Bound    default/www-nfs-web-0   course-nfs-storage            66m
[root@k8smaster1 storageclass]# kubectl get pvc
NAME            STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS         AGE
test-pvc        Bound    pvc-97bce597-0788-49a1-be6d-5a938363797b   1Mi        RWX            course-nfs-storage   113m
test-pvc-sc     Bound    pvc-1d0c10d4-c7f7-433f-8143-78b11fd8fe58   1Mi        RWX            sc2                  9m48s
www-nfs-web-0   Bound    pvc-c234c21b-c3c4-4ffb-a14b-aa47cad7183e   1Gi        RWO            course-nfs-storage   66m
www-nfs-web-1   Bound    pvc-7fdeb85f-481e-48c1-9734-284cce8014fb   1Gi        RWO            course-nfs-storage   65m
www-nfs-web-2   Bound    pvc-7810f38b-2779-49e3-84f2-4b56e16df419   1Gi        RWO            course-nfs-storage   65m

 

补充: 重启后发现nfs 无效,解决办法:

nfs 服务器需要将nfs 服务启动,并且设置为开机自启动;nfs 客户端将nfs-utils 服务启动并且设置为开机自启动。

补充: 设置了默认的storageclass之后,如果不指定会使用默认的

1. yml 如下:

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: test-claim
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 1Mi

2. 创建后查看: 使用的默认的sc

[root@k8smaster01 storageclass]# kubectl apply -f test-default-sc.yml 
persistentvolumeclaim/test-claim created
[root@k8smaster01 storageclass]# kubectl get pvc
NAME         STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS         AGE
test-claim   Bound    pvc-9f784512-e537-468c-9e3c-2b084776a368   1Mi        RWX            course-nfs-storage   4s

补充: 核心概念解释

PV 的全称是:PersistentVolume(持久化卷),是对底层的共享存储的一种抽象,PV 由管理员进行创建和配置,它和具体的底层的共享存储技术的实现方式有关,比如 Ceph、GlusterFS、NFS 等,都是通过插件机制完成与共享存储的对接。

PVC 的全称是:PersistentVolumeClaim(持久化卷声明),PVC 是用户存储的一种声明,PVC 和 Pod 比较类似,Pod 消耗的是节点,PVC 消耗的是 PV 资源,Pod 可以请求 CPU 和内存,而 PVC 可以请求特定的存储空间和访问模式。对

StorageClass,通过 StorageClass 的定义,管理员可以将存储资源定义为某种类型的资源,比如快速存储、慢速存储等,用户根据 StorageClass 的描述就可以非常直观的知道各种存储资源的具体特性了,这样就可以根据应用的特性去申请合适的存储资源了。

 

posted @ 2022-01-22 09:51  QiaoZhi  阅读(1686)  评论(0编辑  收藏  举报