动态存储管理实战：GlusterFS

文件转载自：https://www.orchome.com/1284

本节以GlusterFS为例，从定义StorageClass、创建GlusterFS和Heketi服务、用户申请PVC到创建Pod使用存储资源，对StorageClass和动态资源分配进行详细说明，进一步剖析k8s的存储机制。

1.准备工作

首先在用于部署GlusterFS的三个节点上安装GlusterFS客户端：

yum -y install glusterfs glusterfs-fuse

GlusterFS管理服务容器需要以特权模式运行，kube-apiserver的启动参数中确认已经打开了：

--allow-privileged=true

给要部署GlusterFS管理服务的节点打上"storagenode=glusterfs"的标签，这样可以将GlusterFS容器定向部署到安装了GlusterFS的Node上：

[k8s@kube-server harbor]$ kubectl label node kube-node1 storagenode=glusterfs
node "kube-node1" labeled
[k8s@kube-server harbor]$ kubectl label node kube-node2 storagenode=glusterfs
node "kube-node2" labeled
[k8s@kube-server harbor]$ kubectl label node kube-node3 storagenode=glusterfs
node "kube-node3" labeled

2.创建GlusterFS服务容器集群

GlusterFS服务容器以DaemonSet的方式进行部署，确保每台Node上都运行一个GlusterFS管理服务，glusterfs-daemonset.yaml内容如下。参照 https://github.com/gluster/gluster-kubernetes。

1）在各个Node节点的启动参数中增加以下选项，因为GlusterFS需要使用容器的特权模式运行

--allow-privileged

生效：

systemctl daemon-reload
systemctl restart kubelet
systemctl status kubelet

2）给每个运行GlusterFS的Node节点增加一块数据磁盘

注意数据盘挂载后，在系统中使用的设备描述符，需要在下一步配置中使用到。

3）编辑topology.json拓朴文件

获取一份安装资源：git clone https://github.com/gluster/gluster-kubernetes.git

[k8s@kube-server deploy]$ pwd
/home/k8s/gluster-kubernetes/deploy
[k8s@kube-server deploy]$ ls
gk-deploy  heketi.json.template  kube-templates  ocp-templates  topology.json

至少需要3个节点，按下面格式对该文件进行更新：

[k8s@kube-server deploy]$ cat topology.json
{
  "clusters": [
    {
      "nodes": [
        {
          "node": {
            "hostnames": {
              "manage": [
                "kube-node1"
              ],
              "storage": [
                "172.16.10.101"
              ]
            },
            "zone": 1
          },
          "devices": [
            "/dev/sdb"
          ]
        },
        {
          "node": {
            "hostnames": {
              "manage": [
                "kube-node2"
              ],
              "storage": [
                "172.16.10.102"
              ]
            },
            "zone": 1
          },
          "devices": [
            "/dev/sdb"
          ]
        },
        {
          "node": {
            "hostnames": {
              "manage": [
                "kube-node3"
              ],
              "storage": [
                "172.16.10.103"
              ]
            },
            "zone": 1
          },
          "devices": [
            "/dev/sdb"
          ]
        }
      ]
    }
  ]
}

4）在k8s上部署 GlusterFS + heketi

需要先检查下环境：

• 至少需要3个节点
• 每个节点上至少提供一个裸块存储设备
• 确保以下端口没有被占用：2222,24007, 24008, 49152~49251
• 在系统中加载以下模块： modprobe dm_snapshot && modprobe dm_mirror && modprobe dm_thin_pool
• 安装依赖包：yum -y install glusterfs-fuse

执行部署命令：

注：-g参数表示要创建出一套glusterfs集群服务。

[k8s@kube-server deploy]$ ./gk-deploy -g

如果一切顺利，在结束时会看到下面的输出：

....
service "heketi" created
deployment.extensions "heketi" created
Waiting for heketi pod to start ... OK
Flag --show-all has been deprecated, will be removed in an upcoming release

heketi is now running and accessible via https://172.30.86.3:8080 . To run
administrative commands you can install 'heketi-cli' and use it as follows:

  # heketi-cli -s https://172.30.86.3:8080 --user admin --secret '<ADMIN_KEY>' cluster list

You can find it at https://github.com/heketi/heketi/releases . Alternatively,
use it from within the heketi pod:

  # /opt/k8s/bin/kubectl -n default exec -i heketi-75dcfb7d44-vj9bk -- heketi-cli -s https://localhost:8080 --user admin --secret '<ADMIN_KEY>' cluster list

For dynamic provisioning, create a StorageClass similar to this:

---
apiVersion: storage.k8s.io/v1beta1
kind: StorageClass
metadata:
  name: glusterfs-storage
provisioner: kubernetes.io/glusterfs
parameters:
  resturl: "https://172.30.86.3:8080"

Deployment complete!

查看下都创建出了哪些服务实例：

[k8s@kube-server deploy]$ kubectl get pods -o wide
NAME                        READY     STATUS    RESTARTS   AGE       IP              NODE
glusterfs-88469             1/1       Running   0          2h        172.16.10.102   kube-node2
glusterfs-lwm4n             1/1       Running   0          2h        172.16.10.103   kube-node3
glusterfs-pfgwb             1/1       Running   0          2h        172.16.10.101   kube-node1
heketi-75dcfb7d44-vj9bk     1/1       Running   0          1h        172.30.86.3     kube-node2
my-nginx-86555897f9-2kn92   1/1       Running   2          8h        172.30.49.2     kube-node1
my-nginx-86555897f9-d95t9   1/1       Running   4          2d        172.30.48.2     kube-node3
[k8s@kube-server deploy]$ kubectl get svc -o wide
NAME                       TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)    AGE       SELECTOR
heketi                     ClusterIP   10.254.42.129    <none>        8080/TCP   1h        glusterfs=heketi-pod
heketi-storage-endpoints   ClusterIP   10.254.4.122     <none>        1/TCP      1h        <none>
kubernetes                 ClusterIP   10.254.0.1       <none>        443/TCP    7d        <none>
my-nginx                   ClusterIP   10.254.191.237   <none>        80/TCP     5d        run=my-nginx
[k8s@kube-server deploy]$ kubectl get deployment
NAME       DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
heketi     1         1         1            1           1h
my-nginx   2         2         2            2           5d

[k8s@kube-server deploy]$ kubectl get secret
NAME                                 TYPE                                  DATA      AGE
default-token-p5wjd                  kubernetes.io/service-account-token   3         7d
heketi-config-secret                 Opaque                                3         1h
heketi-service-account-token-mrtsx   kubernetes.io/service-account-token   3         2h
kubelet-api-test-token-gdj7g         kubernetes.io/service-account-token   3         6d
[k8s@kube-server deploy]$

5）使用示例

在可以调用kubectl管理k8s集群的节点上，安装一个heketi客户端：

yum -y install heketi-client

创建个1GB的PV存储卷：

[k8s@kube-server deploy]$ export HEKETI_CLI_SERVER=https://172.30.86.3:8080
[k8s@kube-server deploy]$ heketi-cli volume create --size=1 --persistent-volume --persistent-volume-endpoint=heketi-storage-endpoints | kubectl create -f -
persistentvolume "glusterfs-900fb349" created

在Dashboard上看看这个刚创建的存储卷：

通过heketi服务查看和管理GlusterFS集群： 查看集群列表：

[root@kube-node1 ~]# curl 10.254.42.129:8080/clusters
{"clusters":["ada54ffbeac15a5c9a7521e0c7d2f636"]}

查看集群详情：

[root@kube-node1 ~]# curl 10.254.42.129:8080/clusters/ada54ffbeac15a5c9a7521e0c7d2f636
{"id":"ada54ffbeac15a5c9a7521e0c7d2f636","nodes":["49ac6f56ef21408bcad7c7613cd40bd8","bdf51ae46025cd4fcf134f7be36c32de","fc21262379ec3636e3eadcae15efcc94"],"volumes":["42b01b9b08af23b751b2359fb161c004","900fb349e56af275f47d523d08fdfd6e"],"block":true,"file":true,"blockvolumes":[]}

查看节点详情：

state 为 online说明节点正常

[root@kube-node1 ~]# curl 10.254.42.129:8080/nodes/49ac6f56ef21408bcad7c7613cd40bd8
{"zone":1,"hostnames":{"manage":["kube-node3"],"storage":["172.16.10.103"]},"cluster":"ada54ffbeac15a5c9a7521e0c7d2f636","id":"49ac6f56ef21408bcad7c7613cd40bd8","state":"online","devices":[{"name":"/dev/sdb","storage":{"total":8253440,"free":5087232,"used":3166208},"id":"2f6b2f6c289a2f6bf48fbec59c0c2009","state":"online","bricks":[{"id":"2ea90ebd791a4230e927d233d1c8a7d1","path":"/var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_2ea90ebd791a4230e927d233d1c8a7d1/brick","device":"2f6b2f6c289a2f6bf48fbec59c0c2009","node":"49ac6f56ef21408bcad7c7613cd40bd8","volume":"42b01b9b08af23b751b2359fb161c004","size":2097152},{"id":"4c98684d878ffe7dbfc1008336460eed","path":"/var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_4c98684d878ffe7dbfc1008336460eed/brick","device":"2f6b2f6c289a2f6bf48fbec59c0c2009","node":"49ac6f56ef21408bcad7c7613cd40bd8","volume":"900fb349e56af275f47d523d08fdfd6e","size":1048576}]}]}

6）创建一个使用GlusterFS动态存储供应服务的nginx应用

注：在本示例中的用户认证是未启用的，如果要启动用户认证服务，则可以创建一个secret，然后通过StorageClass配置参数传递给Gluster动态存储供应服务。

下面是一个存储类的示例，它将请求2GB的按需存储，用于在我们的HelloWorld应用程序中使用。

# cat gluster-storage-class.yaml
apiVersion: storage.k8s.io/v1beta1
kind: StorageClass
metadata:
  name: gluster-heketi
provisioner: kubernetes.io/glusterfs
parameters:
  resturl: "https://10.254.42.129:8080"
  restuser: "joe"
  restuserkey: "My Secret Life"

• name，StorageClass名称
• provisioner，存储服务提供者
• resturl，Heketi REST Url
• restuser，因为未启用认证，所以这个参数无效
• restuserkey，同上

创建该存储类：

[k8s@kube-server ~]$ kubectl create -f gluster-storage-class.yaml
storageclass.storage.k8s.io "gluster-heketi" created

[k8s@kube-server ~]$ kubectl get storageclass
NAME             PROVISIONER               AGE
gluster-heketi   kubernetes.io/glusterfs   43s

创建PersistentVolumeClaim（PVC）以请求我们的HelloWorld应用程序的存储：

我们将创建一个要求2GB存储空间的PVC，此时，Kubernetes Dynamic Provisioning Framework和Heketi将自动配置新的GlusterFS卷并生成Kubernetes PersistentVolume（PV）对象。

annotations，Kubernetes存储类注释和存储类的名称

gluster-pvc.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: gluster1
  annotations:
    volume.beta.kubernetes.io/storage-class: gluster-heketi
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi

[k8s@kube-server ~]$ kubectl create -f gluster-pvc.yaml
persistentvolumeclaim "gluster1" created

可以看到PVC是绑定到一个动态供给的存储卷上的：

[k8s@kube-server ~]$ kubectl get pvc
NAME       STATUS    VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS     AGE
gluster1   Bound     pvc-53e824cf-7eb7-11e8-bf5c-080027395360   2Gi        RWO            gluster-heketi   53s

[k8s@kube-server ~]$ kubectl get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM              STORAGECLASS     REASON    AGE
glusterfs-900fb349                         1Gi        RWX            Retain           Available                                                 2h
pvc-53e824cf-7eb7-11e8-bf5c-080027395360   2Gi        RWO            Delete           Bound       default/gluster1   gluster-heketi             1m

创建一个使用该PVC的nginx实例：

nginx-pod.yaml
apiVersion: v1
kind: Pod
metadata:
  name: nginx-pod1
  labels:
    name: nginx-pod1
spec:
  containers:
  - name: nginx-pod1
    image: nginx:1.7.9
    ports:
    - name: web
      containerPort: 80
    volumeMounts:
    - name: gluster-vol1
      mountPath: /usr/share/nginx/html
  volumes:
  - name: gluster-vol1
    persistentVolumeClaim:
      claimName: gluster1 # claimName，要使用的PVC的名称

[k8s@kube-server ~]$ kubectl create -f nginx-pod.yaml
pod "nginx-pod1" created

[k8s@kube-server ~]$ kubectl get pods -o wide|grep nginx-pod
nginx-pod1                  1/1       Running   0          33s       172.30.86.3     kube-node2

登录到该Pod中并创建一个网页文件：

[k8s@kube-server ~]$ kubectl exec -it nginx-pod1  /bin/bash
root@nginx-pod1:/# df -h
Filesystem                                          Size  Used Avail Use% Mounted on
rootfs                                               41G  7.1G   34G  18% /
overlay                                              41G  7.1G   34G  18% /
tmpfs                                                64M     0   64M   0% /dev
tmpfs                                               496M     0  496M   0% /sys/fs/cgroup
/dev/mapper/centos_bogon-root                        41G  7.1G   34G  18% /dev/termination-log
shm                                                  64M     0   64M   0% /dev/shm
/dev/mapper/centos_bogon-root                        41G  7.1G   34G  18% /etc/resolv.conf
/dev/mapper/centos_bogon-root                        41G  7.1G   34G  18% /etc/hostname
/dev/mapper/centos_bogon-root                        41G  7.1G   34G  18% /etc/hosts
/dev/mapper/centos_bogon-root                        41G  7.1G   34G  18% /var/cache/nginx
172.16.10.101:vol_1b6e32efd9b6f07e2b056bed2ce6cc73  2.0G   53M  2.0G   3% /usr/share/nginx/html
tmpfs                                               496M   12K  496M   1% /run/secrets/kubernetes.io/serviceaccount
tmpfs                                                64M     0   64M   0% /proc/kcore
tmpfs                                                64M     0   64M   0% /proc/keys
tmpfs                                                64M     0   64M   0% /proc/timer_list
tmpfs                                                64M     0   64M   0% /proc/timer_stats
tmpfs                                                64M     0   64M   0% /proc/sched_debug
tmpfs                                               496M     0  496M   0% /proc/scsi
tmpfs                                               496M     0  496M   0% /sys/firmware

root@nginx-pod1:/# cd /usr/share/nginx/html
dex.htmlnx-pod1:/usr/share/nginx/html# echo 'Hello World from GlusterFS!!!' > in
root@nginx-pod1:/usr/share/nginx/html# ls
index.html
root@nginx-pod1:/usr/share/nginx/html# cat index.html
Hello World from GlusterFS!!!
root@nginx-pod1:/usr/share/nginx/html# exit
exit

访问一下我们的网页：

[k8s@kube-server ~]$ curl https://172.30.86.3
Hello World from GlusterFS!!!

再检查一下gluster pod，找到我们刚写入的index.html文件，登录任一个gluster pod：

[root@kube-node1 brick]# pwd
/var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_6e016b1ed8e16b6a28839f1670a56d00/brick
[root@kube-node1 brick]# ls
index.html
[root@kube-node1 brick]# cat index.html
Hello World from GlusterFS!!!

7）删除 glusterfs 集群配置

curl -X DELETE 10.254.42.129:8080/devices/46b2685901f56d6fe0cc85bf3d37bf75 # 使用的是device id，删除device
curl -X DELETE 10.254.42.129:8080/nodes/8bd8497a8dcda0708508228f4ae8c2ae #使用的是node id， 删除node

每个节点都要删除掉device才能再删除node

cluster 列表下的所有节点都删除后 才能删除cluster：
curl -X DELETE 10.254.42.129:8080/clusters/ada54ffbeac15a5c9a7521e0c7d2f636

Heketi服务

GlusterFS 是个开源的分布式文件系统，而 Heketi 在其上提供了 REST 形式的 API，二者协同为 Kubernetes 提供了存储卷的自动供给能力。Heketi还支持GlusterFS多集群管理。当一个集群中同时有多种规格、性能和容量特点的存储资源时，Heketi可以通过接入多个存储集群，在每个集群中又按资源特性划分出多个Zone来进行管理。

1）在k8s中部署Heketi服务前需要为其创建一个ServiceAccount账号

我们继续看一下前面例子中使用到的一些配置资源：

[k8s@kube-server deploy]$ pwd
/home/k8s/gluster-kubernetes/deploy
[k8s@kube-server deploy]$ cat ./kube-templates/heketi-service-account.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
  name: heketi-service-account
  labels:
    glusterfs: heketi-sa
    heketi: sa

[k8s@kube-server deploy]$ kubectl get sa | grep heketi
heketi-service-account   1         4h

2）通过Deployment部署Heketi服务

[k8s@kube-server deploy]$ kubectl get deployment
NAME       DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
heketi     1         1         1            1           3h

[k8s@kube-server kube-templates]$ pwd
/home/k8s/gluster-kubernetes/deploy/kube-templates
[k8s@kube-server kube-templates]$ cat heketi-deployment.yaml
---
kind: Service
apiVersion: v1
metadata:
  name: heketi
  labels:
    glusterfs: heketi-service
    heketi: service
  annotations:
    description: Exposes Heketi Service
spec:
  selector:
    glusterfs: heketi-pod
  ports:
  - name: heketi
    port: 8080
    targetPort: 8080
---
kind: Deployment
apiVersion: extensions/v1beta1
metadata:
  name: heketi
  labels:
    glusterfs: heketi-deployment
    heketi: deployment
  annotations:
    description: Defines how to deploy Heketi
spec:
  replicas: 1
  template:
    metadata:
      name: heketi
      labels:
        glusterfs: heketi-pod
        heketi: pod
    spec:
      serviceAccountName: heketi-service-account
      containers:
      - image: heketi/heketi:dev
        imagePullPolicy: IfNotPresent
        name: heketi
        env:
        - name: HEKETI_USER_KEY
          value: ${HEKETI_USER_KEY}
        - name: HEKETI_ADMIN_KEY
          value: ${HEKETI_ADMIN_KEY}
        - name: HEKETI_EXECUTOR
          value: ${HEKETI_EXECUTOR}
        - name: HEKETI_FSTAB
          value: ${HEKETI_FSTAB}
        - name: HEKETI_SNAPSHOT_LIMIT
          value: '14'
        - name: HEKETI_KUBE_GLUSTER_DAEMONSET
          value: "y"
        - name: HEKETI_IGNORE_STALE_OPERATIONS
          value: "true"
        ports:
        - containerPort: 8080
        volumeMounts:
        - name: db
          mountPath: "/var/lib/heketi"
        - name: config
          mountPath: /etc/heketi
        readinessProbe:
          timeoutSeconds: 3
          initialDelaySeconds: 3
          httpGet:
            path: "/hello"
            port: 8080
        livenessProbe:
          timeoutSeconds: 3
          initialDelaySeconds: 30
          httpGet:
            path: "/hello"
            port: 8080
      volumes:
      - name: db
        glusterfs:
          endpoints: heketi-storage-endpoints
          path: heketidbstorage
      - name: config
        secret:
          secretName: heketi-config-secret

定义了两个volumes，其中"db"是使用的glusterfs提供的存储卷，而"config"的volume则是使用的"secret"

查看一下被当作volume使用的secret的内容：

[k8s@kube-server kube-templates]$ kubectl describe secret heketi-config-secret
Name:         heketi-config-secret
Namespace:    default
Labels:       glusterfs=heketi-config-secret
              heketi=config-secret
Annotations:  <none>

Type:  Opaque

Data
====
heketi.json:    909 bytes
private_key:    0 bytes
topology.json:  1029 bytes

可以看到这个secret中是包含了三个配置文件

这个secret是使用脚本创建的：

[k8s@kube-server deploy]$ pwd
/home/k8s/gluster-kubernetes/deploy
[k8s@kube-server deploy]$ ls
gk-deploy  heketi.json.template  kube-templates  ocp-templates  topology.json
[k8s@kube-server deploy]$ grep heketi-config-secret gk-deploy
  eval_output "${CLI} create secret generic heketi-config-secret --from-file=private_key=${SSH_KEYFILE} --from-file=./heketi.json --from-file=topology.json=${TOPOLOGY}"
  eval_output "${CLI} label --overwrite secret heketi-config-secret glusterfs=heketi-config-secret heketi=config-secret"
[k8s@kube-server deploy]$

3）为Heketi设置GlusterFS集群

使用一个topology.json的配置文件来完成各个GlusterFS节点和设备的定义。

Heketi要求一个GlusterFS集群中至少有3个节点。在topology.json配置文件的hostnames字段中的manage中填写主机名，在storage上填写IP地址，devices要求是未创建文件系统的裸设备（支持多块磁盘）。这样Heketi就可以自动完成PV、VG和LV的创建。

topology.json文件的内容在前面已经提供了，不再重复。

当使用gluster-kubernetes项目提供的脚本工具和模板创建glusterfs和heketi服务时，不需要额外的手动干预。但如果是手动创建这些服务，则可以按下面的方法使用Heketi加载topology配置，完成GlusterFS服务集群的创建。

可以登录进入Hekiti容器执行以下命令：

# export HEKETI_CLI_SERVER=https://localhost:8080
# heketi-cli topology load --json=topology.json

这样Heketi就完成了GlusterFS集群的创建，同时在GlusterFS集群的各个节点的/dev/sdb盘上成功创建了PV和VG。

此时，查看Heketi的topology信息，可以看到很详细的Node和Device信息：

[k8s@kube-server deploy]$ heketi-cli topology info

Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636

    Volumes:

    Name: vol_1b6e32efd9b6f07e2b056bed2ce6cc73
    Size: 2
    Id: 1b6e32efd9b6f07e2b056bed2ce6cc73
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Mount: 172.16.10.103:vol_1b6e32efd9b6f07e2b056bed2ce6cc73
    Mount Options: backup-volfile-servers=172.16.10.102,172.16.10.101
    Durability Type: replicate
    Replica: 3
    Snapshot: Disabled

        Bricks:
            Id: 6e016b1ed8e16b6a28839f1670a56d00
            Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_6e016b1ed8e16b6a28839f1670a56d00/brick
            Size (GiB): 2
            Node: fc21262379ec3636e3eadcae15efcc94
            Device: f8776d0d92102fc3e272f2ec899e5f18

            Id: c7acfcecaf85aada98b0c0208798440f
            Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_c7acfcecaf85aada98b0c0208798440f/brick
            Size (GiB): 2
            Node: 49ac6f56ef21408bcad7c7613cd40bd8
            Device: 2f6b2f6c289a2f6bf48fbec59c0c2009

            Id: e2acc8268f17f05e940ebe679dacfa3a
            Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_e2acc8268f17f05e940ebe679dacfa3a/brick
            Size (GiB): 2
            Node: bdf51ae46025cd4fcf134f7be36c32de
            Device: e100120226d5d9567ed0f92b9810236c

    Name: heketidbstorage
    Size: 2
    Id: 42b01b9b08af23b751b2359fb161c004
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Mount: 172.16.10.103:heketidbstorage
    Mount Options: backup-volfile-servers=172.16.10.102,172.16.10.101
    Durability Type: replicate
    Replica: 3
    Snapshot: Disabled

        Bricks:
            Id: 2ea90ebd791a4230e927d233d1c8a7d1
            Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_2ea90ebd791a4230e927d233d1c8a7d1/brick
            Size (GiB): 2
            Node: 49ac6f56ef21408bcad7c7613cd40bd8
            Device: 2f6b2f6c289a2f6bf48fbec59c0c2009

            Id: 9dc7238db3240146f12189dd28320227
            Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_9dc7238db3240146f12189dd28320227/brick
            Size (GiB): 2
            Node: fc21262379ec3636e3eadcae15efcc94
            Device: f8776d0d92102fc3e272f2ec899e5f18

            Id: cb68cbf4e992abffc86ab3b5db58ef56
            Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_cb68cbf4e992abffc86ab3b5db58ef56/brick
            Size (GiB): 2
            Node: bdf51ae46025cd4fcf134f7be36c32de
            Device: e100120226d5d9567ed0f92b9810236c

    Name: vol_900fb349e56af275f47d523d08fdfd6e
    Size: 1
    Id: 900fb349e56af275f47d523d08fdfd6e
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Mount: 172.16.10.103:vol_900fb349e56af275f47d523d08fdfd6e
    Mount Options: backup-volfile-servers=172.16.10.102,172.16.10.101
    Durability Type: replicate
    Replica: 3
    Snapshot: Disabled

        Bricks:
            Id: 4c98684d878ffe7dbfc1008336460eed
            Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_4c98684d878ffe7dbfc1008336460eed/brick
            Size (GiB): 1
            Node: 49ac6f56ef21408bcad7c7613cd40bd8
            Device: 2f6b2f6c289a2f6bf48fbec59c0c2009

            Id: 7c82d03c88d73bb18d407a791a1053c2
            Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_7c82d03c88d73bb18d407a791a1053c2/brick
            Size (GiB): 1
            Node: bdf51ae46025cd4fcf134f7be36c32de
            Device: e100120226d5d9567ed0f92b9810236c

            Id: 822266f7ad3cf62b1e45686265cf7268
            Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_822266f7ad3cf62b1e45686265cf7268/brick
            Size (GiB): 1
            Node: fc21262379ec3636e3eadcae15efcc94
            Device: f8776d0d92102fc3e272f2ec899e5f18

    Nodes:

    Node Id: 49ac6f56ef21408bcad7c7613cd40bd8
    State: online
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Zone: 1
    Management Hostname: kube-node3
    Storage Hostname: 172.16.10.103
    Devices:
        Id:2f6b2f6c289a2f6bf48fbec59c0c2009   Name:/dev/sdb            State:online    Size (GiB):7       Used (GiB):5       Free (GiB):2
            Bricks:
                Id:2ea90ebd791a4230e927d233d1c8a7d1   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_2ea90ebd791a4230e927d233d1c8a7d1/brick
                Id:4c98684d878ffe7dbfc1008336460eed   Size (GiB):1       Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_4c98684d878ffe7dbfc1008336460eed/brick
                Id:c7acfcecaf85aada98b0c0208798440f   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_2f6b2f6c289a2f6bf48fbec59c0c2009/brick_c7acfcecaf85aada98b0c0208798440f/brick

    Node Id: bdf51ae46025cd4fcf134f7be36c32de
    State: online
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Zone: 1
    Management Hostname: kube-node2
    Storage Hostname: 172.16.10.102
    Devices:
        Id:e100120226d5d9567ed0f92b9810236c   Name:/dev/sdb            State:online    Size (GiB):7       Used (GiB):5       Free (GiB):2
            Bricks:
                Id:7c82d03c88d73bb18d407a791a1053c2   Size (GiB):1       Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_7c82d03c88d73bb18d407a791a1053c2/brick
                Id:cb68cbf4e992abffc86ab3b5db58ef56   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_cb68cbf4e992abffc86ab3b5db58ef56/brick
                Id:e2acc8268f17f05e940ebe679dacfa3a   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_e100120226d5d9567ed0f92b9810236c/brick_e2acc8268f17f05e940ebe679dacfa3a/brick

    Node Id: fc21262379ec3636e3eadcae15efcc94
    State: online
    Cluster Id: ada54ffbeac15a5c9a7521e0c7d2f636
    Zone: 1
    Management Hostname: kube-node1
    Storage Hostname: 172.16.10.101
    Devices:
        Id:f8776d0d92102fc3e272f2ec899e5f18   Name:/dev/sdb            State:online    Size (GiB):7       Used (GiB):5       Free (GiB):2
            Bricks:
                Id:6e016b1ed8e16b6a28839f1670a56d00   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_6e016b1ed8e16b6a28839f1670a56d00/brick
                Id:822266f7ad3cf62b1e45686265cf7268   Size (GiB):1       Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_822266f7ad3cf62b1e45686265cf7268/brick
                Id:9dc7238db3240146f12189dd28320227   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_f8776d0d92102fc3e272f2ec899e5f18/brick_9dc7238db3240146f12189dd28320227/brick

总结：使用Kubernetes的动态存储供应模式，配置StorageClass和Heketi共同搭建基于GlusterFS的共享存储，相对于静态模式至少有两大优势。

1.一个是管理员无须预先创建大量的PV作为存储资源

2.用户在申请PVC时也无法保证容量与预置PV的容量能够一致。因此，从k8s v1.6开始，建议用户优先考虑使用StorageClass的动态存储供应模式进行存储管理。