k8s实践12:kubernetes对接nfs存储实现pvc动态按需创建分配绑定pv

 

am2012关注1人评论2390人阅读2019-07-24 16:34:26

 

1.
开始前的想法.
前面测试pv&&pvc的部署和简单配置应用,实现pod应用数据存储到pvc并且和pod解耦的目的.
前面操作是全手动操作,手动创建pv,手动创建pvc,如果集群pod少,这样操作可以.
假如集群有1000个以上的pod,每个pod都需要使用pvc存储数据,如果只能手动去一个个创建pv,pvc,工作量不可想像.
如果可以创建pod的时候,创建pod的用户定义pvc,然后集群能够根据用户的pvc需求创建pv,实现动态的pv&&pvc创建分配.
kubernetes支持对接存储动态创建分配pv&&pvc.
这是本次测试的目的.

2.
测试环境

实验环境,存储用nfs简单部署测试.

3.
nfs部署

略
参考前面的文档
pod应用数据存储解耦pv&&pvc

4.
storage classes

官方文档:
https://kubernetes.io/docs/concepts/storage/storage-classes/
kubernetes支持用storage classes对接存储,实现动态pv&&pvc创建分配.
kubernetes内置支持对接很多存储类型,比如cephfs,glusterfs等等,具体参考官方文档.
kubernetes内置不支持对接nfs存储类型.需要使用外部的插件.
外部插件参考文档:
https://github.com/kubernetes-incubator/external-storage
nfs插件配置文档:
https://github.com/kubernetes-incubator/external-storage/tree/master/nfs-client
nfs-client-provisioner是一个kubernetes的简易NFS的外部provisioner,本身不提供NFS,需要现有的NFS服务器提供存储

5.
nfs存储配置文件

[root@k8s-master1 nfs]# ls
class.yaml  deployment.yaml  rbac.yaml  test-claim.yaml  test-pod.yaml

5.1
class.yaml

[root@k8s-master1 nfs]# cat class.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: managed-nfs-storage
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
parameters:
  archiveOnDelete: "false"

创建一个storageclass
kind: StorageClass

新建的storageclass名字为:managed-nfs-storage                                         
name: managed-nfs-storage
 
provisioner直译为供应者,结合实际这里应该是指storageclass的对接存储类程序名字(个人理解),这个名字必须和deplotment.yaml的PROVISIONER_NAME变量值相同.                      
provisioner: fuseim.pri/ifs                                         

[root@k8s-master1 nfs]# kubectl apply -f class.yaml
storageclass.storage.k8s.io "managed-nfs-storage" created

[root@k8s-master1 nfs]# kubectl get storageclass
NAME                  PROVISIONER      AGE
managed-nfs-storage   fuseim.pri/ifs   7s

5.2
deployment.yaml

[root@k8s-master1 nfs]# cat deployment.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
  name: nfs-client-provisioner
---
kind: Deployment
apiVersion: extensions/v1beta1
metadata:
  name: nfs-client-provisioner
spec:
  replicas: 1
  strategy:
    type: Recreate
  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: 10.10.10.60
            - name: NFS_PATH
              value: /ifs/kubernetes
      volumes:
        - name: nfs-client-root
          nfs:
            server: 10.10.10.60
            path: /ifs/kubernetes
[root@k8s-master1 nfs]#

创建sa,名字为:nfs-client-provisioner

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

pod名字和使用的镜像

containers:
        - name: nfs-client-provisioner
          image: quay.io/external_storage/nfs-client-provisioner:latest

pod里挂载的路径

 volumeMounts:
            - name: nfs-client-root
              mountPath: /persistentvolumes

pod读取的变量,这里需要修改成本地nfs的地址和路径

 env:
            - name: PROVISIONER_NAME
              value: fuseim.pri/ifs
            - name: NFS_SERVER
              value: 10.10.10.60
            - name: NFS_PATH
              value: /ifs/kubernetes

nfs服务的地址和路径,需要修改成本地nfs的地址和路径

 volumes:
        - name: nfs-client-root
          nfs:
            server: 10.10.10.60
            path: /ifs/kubernetes 

修改后的deployment.yaml文件,只是修改了nfs的地址和目录

[root@k8s-master1 nfs]# cat deployment.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
  name: nfs-client-provisioner
---
kind: Deployment
apiVersion: extensions/v1beta1
metadata:
  name: nfs-client-provisioner
spec:
  replicas: 1
  strategy:
    type: Recreate
  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: 192.168.32.130
            - name: NFS_PATH
              value: /mnt/k8s
      volumes:
        - name: nfs-client-root
          nfs:
            server: 192.168.32.130
            path: /mnt/k8s

[root@k8s-master1 nfs]# kubectl apply -f deployment.yaml
serviceaccount "nfs-client-provisioner" created
deployment.extensions "nfs-client-provisioner" created
[root@k8s-master1 nfs]#

[root@k8s-master1 nfs]# kubectl get pod
NAME                                      READY     STATUS    RESTARTS   AGE
nfs-client-provisioner-65bf6bd464-qdzcj   1/1       Running   0          1m

[root@k8s-master1 nfs]# kubectl describe pod nfs-client-provisioner-65bf6bd464-qdzcj
Name:               nfs-client-provisioner-65bf6bd464-qdzcj
Namespace:          default
Priority:           0
PriorityClassName:  <none>
Node:               k8s-master3/192.168.32.130
Start Time:         Wed, 24 Jul 2019 14:44:11 +0800
Labels:             app=nfs-client-provisioner
                    pod-template-hash=65bf6bd464
Annotations:        <none>
Status:             Running
IP:                 172.30.35.3
Controlled By:      ReplicaSet/nfs-client-provisioner-65bf6bd464
Containers:
  nfs-client-provisioner:
    Container ID:   docker://67329cd9ca608223cda961a1bfe11524f2586e8e1ccba45ad57b292b1508b575
    Image:          quay.io/external_storage/nfs-client-provisioner:latest
    Image ID:       docker-pullable://quay.io/external_storage/nfs-client-provisioner@sha256:022ea0b0d69834b652a4c53655d78642ae23f0324309097be874fb58d09d2919
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Wed, 24 Jul 2019 14:45:52 +0800
    Ready:          True
    Restart Count:  0
    Environment:
      PROVISIONER_NAME:  fuseim.pri/ifs
      NFS_SERVER:        192.168.32.130
      NFS_PATH:          /mnt/k8s
    Mounts:
      /persistentvolumes from nfs-client-root (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from nfs-client-provisioner-token-4n4jn (ro)
Conditions:
  Type              Status
  Initialized       True
  Ready             True
  ContainersReady   True
  PodScheduled      True
Volumes:
  nfs-client-root:
    Type:      NFS (an NFS mount that lasts the lifetime of a pod)
    Server:    192.168.32.130
    Path:      /mnt/k8s
    ReadOnly:  false
  nfs-client-provisioner-token-4n4jn:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  nfs-client-provisioner-token-4n4jn
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type    Reason     Age   From                  Message
  ----    ------     ----  ----                  -------
  Normal  Scheduled  2m    default-scheduler     Successfully assigned default/nfs-client-provisioner-65bf6bd464-qdzcj to k8s-master3
  Normal  Pulling    2m    kubelet, k8s-master3  pulling image "quay.io/external_storage/nfs-client-provisioner:latest"
  Normal  Pulled     54s   kubelet, k8s-master3  Successfully pulled image "quay.io/external_storage/nfs-client-provisioner:latest"
  Normal  Created    54s   kubelet, k8s-master3  Created container
  Normal  Started    54s   kubelet, k8s-master3  Started container
[root@k8s-master1 nfs]#

5.3
rbac.yaml
指定sa:nfs-client-provisioner的权限
nfs-client-provisioner在deployment部署时,已经创建.

[root@k8s-master1 nfs]# cat rbac.yaml
kind: ServiceAccount
apiVersion: v1
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: ["create", "update", "patch"]
---
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
---
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: leader-locking-nfs-client-provisioner
rules:
  - apiGroups: [""]
    resources: ["endpoints"]
    verbs: ["get", "list", "watch", "create", "update", "patch"]
---
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: leader-locking-nfs-client-provisioner
subjects:
  - kind: ServiceAccount
    name: nfs-client-provisioner
    # replace with namespace where provisioner is deployed
    namespace: default
roleRef:
  kind: Role
  name: leader-locking-nfs-client-provisioner
  apiGroup: rbac.authorization.k8s.io
[root@k8s-master1 nfs]#

[root@k8s-master1 nfs]# kubectl apply -f rbac.yaml
serviceaccount "nfs-client-provisioner" unchanged
clusterrole.rbac.authorization.k8s.io "nfs-client-provisioner-runner" created
clusterrolebinding.rbac.authorization.k8s.io "run-nfs-client-provisioner" created
role.rbac.authorization.k8s.io "leader-locking-nfs-client-provisioner" created
rolebinding.rbac.authorization.k8s.io "leader-locking-nfs-client-provisioner" created
[root@k8s-master1 nfs]#

检索下

[root@k8s-master1 nfs]# kubectl get clusterrole |grep nfs
nfs-client-provisioner-runner                                          2m
[root@k8s-master1 nfs]# kubectl get role |grep nfs
leader-locking-nfs-client-provisioner   2m
[root@k8s-master1 nfs]# kubectl get rolebinding |grep nfs
leader-locking-nfs-client-provisioner   2m
[root@k8s-master1 nfs]# kubectl get clusterrolebinding |grep nfs
run-nfs-client-provisioner                             2m
[root@k8s-master1 nfs]#

6.
测试

使用官方的test-claim.yaml测试

[root@k8s-master1 nfs]# cat test-claim.yaml
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: test-claim
  annotations:
    volume.beta.kubernetes.io/storage-class: "managed-nfs-storage"
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 1Mi

读取执行test.claim.yaml文件的pv,pvc情况

[root@k8s-master1 nfs]# kubectl get pv
No resources found.
[root@k8s-master1 nfs]# kubectl get pvc
No resources found.
[root@k8s-master1 nfs]#

读取执行

[root@k8s-master1 nfs]# kubectl apply -f test-claim.yaml
persistentvolumeclaim "test-claim" created

执行后的pv,pvc情况

[root@k8s-master1 nfs]# kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS    CLAIM                STORAGECLASS          REASON    AGE
pvc-4fb682ac-ade0-11e9-8401-000c29383c89   1Mi        RWX            Delete           Bound     default/test-claim   managed-nfs-storage             6s
[root@k8s-master1 nfs]# kubectl get pvc
NAME         STATUS    VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS          AGE
test-claim   Bound     pvc-4fb682ac-ade0-11e9-8401-000c29383c89   1Mi        RWX            managed-nfs-storage   8s
[root@k8s-master1 nfs]#

成功了.对接nfs存储类后,用户可以申请创建pvc,系统自动创建pv并绑定pvc.
检索nfs server的存储目录

[root@k8s-master3 k8s]# pwd
/mnt/k8s
[root@k8s-master3 k8s]# ls
default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89
[root@k8s-master3 k8s]#

检索pod里的挂载目录

[root@k8s-master1 nfs]# kubectl exec -it nfs-client-provisioner-65bf6bd464-qdzcj ls /persistentvolumes
default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89
[root@k8s-master1 nfs]#

7.
使用官方的test-pod.yaml测试

[root@k8s-master1 nfs]# cat test-pod.yaml
kind: Pod
apiVersion: v1
metadata:
  name: test-pod
spec:
  containers:
  - name: test-pod
    image: gcr.io/google_containers/busybox:1.24
    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-claim
[root@k8s-master1 nfs]#

[root@k8s-master1 nfs]# kubectl apply -f test-pod.yaml
pod "test-pod" created

[root@k8s-master1 nfs]# kubectl get pod
NAME                                      READY     STATUS      RESTARTS   AGE
test-pod                                  0/1       Completed   0          1m

pod启动后,在/mnt目录创建了文件SUCCESS
pvc挂载的pod目录就是/mnt
在nfs server目录可以看到test-pod创建的SUCCESS文件:

[root@k8s-master3 default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89]# pwd
/mnt/k8s/default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89
[root@k8s-master3 default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89]# ls
SUCCESS

检索nfs-client-provisioner

[root@k8s-master1 nfs]# kubectl exec -it nfs-client-provisioner-65bf6bd464-qdzcj ls /persistentvolumes/default-test-claim-pvc-4fb682ac-ade0-11e9-8401-000c29383c89
SUCCESS

8.
测试之后的一个疑问

删除pod,pvc存储的数据还在,删除pvc之后,pvc目录和存储的数据都丢失.
为了防止用户操作失误,是否可以保留一份备份呢?
答案是可以.

[root@k8s-master1 nfs]# cat class.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: managed-nfs-storage
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
parameters:
  archiveOnDelete: "false"
[root@k8s-master1 nfs]# cat class.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: managed-nfs-storage
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
parameters:
  archiveOnDelete: "false"

archiveOnDelete: "false"   
这个参数可以设置为false和true.
archiveOnDelete字面意思为删除时是否存档,false表示不存档,即删除数据,true表示存档,即重命名路径.

修改测试

[root@k8s-master1 nfs]# kubectl get storageclass
NAME                  PROVISIONER      AGE
managed-nfs-storage  fuseim.pri/ifs  1m
[root@k8s-master1 nfs]# kubectl describe storageclass
Name:            managed-nfs-storage
IsDefaultClass:  No
Annotations:    kubectl.kubernetes.io/last-applied-configuration={"apiVersion":"storage.k8s.io/v1","kind":"StorageClass","metadata":{"annotations":{},"name":"managed-nfs-storage","namespace":""},"parameters":{"archiveOnDelete":"true"},"provisioner":"fuseim.pri/ifs"}

Provisioner:          fuseim.pri/ifs
Parameters:            archiveOnDelete=true
AllowVolumeExpansion:  <unset>
MountOptions:          <none>
ReclaimPolicy:        Delete
VolumeBindingMode:    Immediate
Events:                <none>

删除pod,pvc

[root@k8s-master1 nfs]# kubectl get pod
NAME                                      READY    STATUS      RESTARTS  AGE
test-pod                                  0/1      Completed  0          6s
[root@k8s-master1 nfs]# kubectl get pv,pvc
NAME                                                        CAPACITY  ACCESS MODES  RECLAIM POLICY  STATUS    CLAIM                STORAGECLASS          REASON    AGE
persistentvolume/pvc-5a12cb0e-adeb-11e9-8401-000c29383c89  1Mi        RWX            Delete          Bound    default/test-claim  managed-nfs-storage            17s

NAME                              STATUS    VOLUME                                    CAPACITY  ACCESS MODES  STORAGECLASS          AGE
persistentvolumeclaim/test-claim  Bound    pvc-5a12cb0e-adeb-11e9-8401-000c29383c89  1Mi        RWX            managed-nfs-storage  17s

[root@k8s-master1 nfs]# kubectl delete -f test-pod.yaml
pod "test-pod" deleted
[root@k8s-master1 nfs]# kubectl delete -f test-claim.yaml
persistentvolumeclaim "test-claim" deleted

[root@k8s-master1 nfs]# kubectl get pv,pvc
No resources found.
[root@k8s-master1 nfs]#

检索nfs server 存储路径,文件自动做了备份.

 [root@k8s-master3 archived-default-test-claim-pvc-5a12cb0e-adeb-11e9-8401-000c29383c89]# pwd
/mnt/k8s/archived-default-test-claim-pvc-5a12cb0e-adeb-11e9-8401-000c29383c89
[root@k8s-master3 archived-default-test-claim-pvc-5a12cb0e-adeb-11e9-8401-000c29383c89]# ls
SUCCESS

切记用上archiveOnDelete:true

9.
部署nfs存储之后,用户可以自行申请pvc.
不再需要再一个个手动创建pv对应pvc的申请.
其实还是有点不方便,可以不可以创建pod的时候就自动申请创建pvc,而不再需要再创建pod前先申请pvc然后再挂载进pod呢?
这是statefulset里的volumeClaimTemplates的功能.
下篇再来测试.