Kubernetes-动态PV&PVC

目录

PV&PVC基本使用

StorageClass&Provisioner

PVC的扩容

---

PV&PVC基本使用

Kubernetes中PV和PVC的概念就不在介绍了

PV和PVC的使用方式是

管理员先创建一系列各种大小的PV，形成集群中的存储资源池

用户创建pvc时需要指定大小，这个时候会从pv池中寻找合适的大小并将pv绑定到pvc上使用

举例说明kubernetes通过pv/pvc使用glusterfs：

1. 创建glusterfs存储卷，假设已经创建好，我们使用GlusterFS上的afr-volume这个volume。

[root@k8s-node1 json]# gluster volume info afr-volume
  
Volume Name: afr-volume
Type: Replicate
Volume ID: 94fddfe0-123d-4db0-b55b-a49ac2486b21
Status: Started
Snapshot Count: 0
Number of Bricks: 1 x 2 = 2
Transport-type: tcp
Bricks:
Brick1: k8s-node1:/opt/afr_data
Brick2: k8s-node2:/opt/afr_data
Options Reconfigured:
transport.address-family: inet
nfs.disable: on

2. kubernetes中创建glusterfs的endpoint

apiVersion: v1
kind: Endpoints
metadata:
  name: glusterfs-endpoint
  namespace: kube-system
subsets:
- addresses:
  - ip: 172.16.9.201
  - ip: 172.16.9.202
  - ip: 172.16.9.203
  ports:
  - port: 1
    protocol: TCP

3. kubernetes中创建glusterfs service(创建与Endpoint对应的Service，Service的名称必须和Endpoint一致)

apiVersion: v1
kind: Service
metadata:
  name: glusterfs-service
  namespace: kube-system
spec:
  ports:
  - port: 1
    protocol: TCP
    targetPort: 1
  type: ClusterIP

4. 创建pv

apiVersion: v1
kind: PersistentVolume
metadata:
  name: gluster-dev-volume
spec:
  capacity:
    storage: 8Gi
  accessModes:
    - ReadWriteMany
  glusterfs:
    endpoints: "glusterfs-cluster"
    path: "afr-volume"
    readOnly: false

5. 创建pvc

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: glusterfs-nginx
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 8Gi

查看pv和pvc可以看到可PV已经由Availiable变成了Bound状态。

 

StorageClass&Provisioner

但是实际情况下，管理员并不清楚用户需要什么样大小的存储卷，也没有办法在预先创建各种大小的PV

最好的效果是用户创建指定大小的pvc，则就自动创建同样大小的pv并关联用户的pvc

Kubernetes通过创建StorageClass来使用 Dynamic Provisioning 特性，storageclass需要有一个provisioner来决定使用什么样的存储插件来动态创建pvc，比如是glusterfs存储，cephfs存储等等

kubernetes中有很多内置的provisioner，参考如下：

https://kubernetes.io/docs/concepts/storage/storage-classes/#provisioner

 

kubernetes中内置了glusterfs的provisioner：kubernetes.io/glusterfs

glusterfs的provisioner是通过heketi提供的restapi来动态创建pvc的，storageclass如下：

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: replicate-2
parameters:
  clusterid: 888490f4545c9a4cc896a6f7485f0362
  resturl: http://192.168.1.1:8087
  volumetype: replicate:2
provisioner: kubernetes.io/glusterfs
reclaimPolicy: Delete
volumeBindingMode: Immediate

使用storageclass创建pvc就不需要在按部就班的创建 endpoint - service - pv - pvc了

直接创建pvc指定大小即可

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: testpvc
  namespace: kube-system
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 20Gi
  storageClassName: replicate-2

可以看到 provisioner自动创建并关联了相关的资源

[root@k8s-node1 ~]# kubectl --namespace=kube-system get pvc aistack-influxdb -oyaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  annotations:
    pv.kubernetes.io/bind-completed: "yes"
    pv.kubernetes.io/bound-by-controller: "yes"
    volume.beta.kubernetes.io/storage-provisioner: kubernetes.io/glusterfs
  creationTimestamp: "2019-07-02T06:38:13Z"
  finalizers:
  - kubernetes.io/pvc-protection
  name: testpvc
  namespace: kube-system
  resourceVersion: "3192"
  selfLink: /api/v1/namespaces/kube-system/persistentvolumeclaims/testpvc
  uid: f7392a88-9c93-11e9-aa77-88d7f6ae9c94
spec:
  accessModes:
  - ReadWriteMany
  dataSource: null
  resources:
    requests:
      storage: 20Gi
  storageClassName: replicate-2
  volumeMode: Filesystem
  volumeName: pvc-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 20Gi
  phase: Bound

[root@k8s-node1 ~]# kubectl --namespace=kube-system get pv pvc-f7392a88-9c93-11e9-aa77-88d7f6ae9c94 -oyaml
apiVersion: v1
kind: PersistentVolume
metadata:
  annotations:
    Description: 'Gluster-Internal: Dynamically provisioned PV'
    gluster.kubernetes.io/heketi-volume-id: 65314d1b4c5136fef92b7ef1a3725a57
    gluster.org/type: file
    kubernetes.io/createdby: heketi-dynamic-provisioner
    pv.beta.kubernetes.io/gid: "2001"
    pv.kubernetes.io/bound-by-controller: "yes"
    pv.kubernetes.io/provisioned-by: kubernetes.io/glusterfs
  creationTimestamp: "2019-07-02T06:38:25Z"
  finalizers:
  - kubernetes.io/pv-protection
  name: pvc-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
  resourceVersion: "3196"
  selfLink: /api/v1/persistentvolumes/pvc-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
  uid: fe963cc4-9c93-11e9-aa77-88d7f6ae9c94
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 20Gi
  claimRef:
    apiVersion: v1
    kind: PersistentVolumeClaim
    name: testpvc
    namespace: kube-system
    resourceVersion: "3168"
    uid: f7392a88-9c93-11e9-aa77-88d7f6ae9c94
  glusterfs:
    endpoints: glusterfs-dynamic-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
    endpointsNamespace: kube-system
    path: vol_65314d1b4c5136fef92b7ef1a3725a57
  persistentVolumeReclaimPolicy: Delete
  storageClassName: replicate-2
  volumeMode: Filesystem
status:
  phase: Bound


[root@k8s-node1 ~]# kubectl --namespace=kube-system get endpoints glusterfs-dynamic-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
NAME                                                     ENDPOINTS                       AGE
glusterfs-dynamic-f7392a88-9c93-11e9-aa77-88d7f6ae9c94   172.16.2.100:1,172.16.2.200:1   6d

[root@k8s-node1 ~]# kubectl --namespace=kube-system get services glusterfs-dynamic-f7392a88-9c93-11e9-aa77-88d7f6ae9c94
NAME                                                     TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)   AGE
glusterfs-dynamic-f7392a88-9c93-11e9-aa77-88d7f6ae9c94   ClusterIP   10.10.159.255           1/TCP     6d

 

PVC的扩容

在kubernetes 1.11版本中开始支持pvc创建后的扩容

storageclass的创建

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: glusterfs-sc
parameters:
  clusterid: 075e35a0ce70274b3ba7f158e77edb2c
  resturl: http://172.16.9.201:8087
  volumetype: replicate:3
provisioner: kubernetes.io/glusterfs
reclaimPolicy: Delete
volumeBindingMode: Immediate
allowVolumeExpansion: true  # 重要

可以看到storageclass增加了allowVolumeExpansion字段

先测试下不带该字段时的pvc扩容情况

先创建一个1G的pvc

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  annotations:
    volume.beta.kubernetes.io/storage-provisioner: kubernetes.io/glusterfs
  name: testpvc
spec:
  accessModes:
  - ReadWriteMany
  dataSource: null
  resources:
    requests:
      storage: 1Gi
  storageClassName: replicate-3
  volumeMode: Filesystem

kubectl create -f pvc.yaml
persistentvolumeclaim/testpvc created

将该PVC绑定一个Pod

[root@k8s-node1 test]# cat pod.yaml
apiVersion: v1
kind: Pod
metadata:
  name: testpod
spec:
  containers:
  - image: nginx:latest
    imagePullPolicy: IfNotPresent
    name: nginx
    - mountPath: /root/testpvc
      name: testpvc
  volumes:
  - name: testpvc
    persistentVolumeClaim:
      claimName: testpvc

[root@k8s-node1 test]# kubectl create -f pod.yaml
pod/testpod created

进入容器查看PVC大小，可以看到挂载的目录/root/testpvc确实是1G左右

[root@k8s-node1 test]# kubectl exec -it testpod bash
[root@testpod /]# df -h
Filesystem                                         Size  Used Avail Use% Mounted on
overlay                                            1.1T   29G  1.1T   3% /
tmpfs                                               64M     0   64M   0% /dev
tmpfs                                               32G     0   32G   0% /sys/fs/cgroup
172.16.2.100:vol_088693919fcae3397926ad54f7d77815 1020M   33M  987M   4% /root/testpvc
/dev/mapper/k8s-kubelet                            704G   33M  704G   1% /etc/hosts
/dev/mapper/k8s-docker                             1.1T   29G  1.1T   3% /etc/hostname
shm                                                 64M     0   64M   0% /dev/shm
tmpfs                                               32G   12K   32G   1% /run/secrets/kubernetes.io/serviceaccount
tmpfs                                               32G     0   32G   0% /proc/acpi
tmpfs                                               32G     0   32G   0% /proc/scsi
tmpfs                                               32G     0   32G   0% /sys/firmware

修改pvc，将大小修改为2G

[root@k8s-node1 test]# kubectl edit pvc testpvc
...

error: persistentvolumeclaims "testpvc" could not be patched: persistentvolumeclaims "testpvc" is forbidden: only dynamically provisioned pvc can be resized and the storageclass that provisions the pvc must support resize
You can run `kubectl replace -f /tmp/kubectl-edit-vqpx5.yaml` to try this update again.

直接报错，显示为无法修改，只有动态pvc能resize，并且provision要能支持resize

当前环境使用的是glusterfs，修改storageclass，添加allowVolumeExpansion: true

[root@k8s-node1 test]# kubectl --namespace=kube-system edit storageclasses.storage.k8s.io replicate-2
storageclass.storage.k8s.io/replicate-2 edited

[root@k8s-node1 test]# kubectl --namespace=kube-system get storageclasses.storage.k8s.io replicate-2 -oyaml
allowVolumeExpansion: true
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  creationTimestamp: "2019-07-02T06:08:00Z"
  name: replicate-2
  resourceVersion: "1261960"
  selfLink: /apis/storage.k8s.io/v1/storageclasses/replicate-2
  uid: bee4b8f2-9c8f-11e9-aa77-88d7f6ae9c94
parameters:
  clusterid: 888490f4545c9a4cc896a6f7485f0362
  resturl: http://172.16.2.100:8087
  volumetype: replicate:2
provisioner: kubernetes.io/glusterfs
reclaimPolicy: Delete
volumeBindingMode: Immediate

可以看到storageclass中已经添加了allowVolumeExpansion: true字段，再次修改pvc，将1G改成2G，这次没有报错

[root@k8s-node1 test]# kubectl edit pvc testpvc
persistentvolumeclaim/testpvc edited

查看pvc可以看到大小已经变成了2G

[root@k8s-node1 test]# kubectl get pvc testpvc
NAME      STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
testpvc   Bound    pvc-6880b78a-a450-11e9-aa77-88d7f6ae9c94   2Gi        RWX            replicate-2    13m

进入到容器中查看实际的大小，可以看到/root/testpvc目录也变成了2G

[root@testpod /]# df -h
Filesystem                                         Size  Used Avail Use% Mounted on
overlay                                            1.1T   29G  1.1T   3% /
tmpfs                                               64M     0   64M   0% /dev
tmpfs                                               32G     0   32G   0% /sys/fs/cgroup
172.16.2.100:vol_088693919fcae3397926ad54f7d77815  2.0G   66M  2.0G   4% /root/testpvc
/dev/mapper/k8s-kubelet                            704G   33M  704G   1% /etc/hosts
/dev/mapper/k8s-docker                             1.1T   29G  1.1T   3% /etc/hostname
shm                                                 64M     0   64M   0% /dev/shm
tmpfs                                               32G   12K   32G   1% /run/secrets/kubernetes.io/serviceaccount
tmpfs                                               32G     0   32G   0% /proc/acpi
tmpfs                                               32G     0   32G   0% /proc/scsi
tmpfs                                               32G     0   32G   0% /sys/firmware