kubernetes存储之GlusterFS（GlusterFS for Kubernetes Storage）

💝💝💝欢迎来到我的博客，很高兴能够在这里和您见面！希望您在这里可以感受到一份轻松愉快的氛围，不仅可以获得有趣的内容和知识，也可以畅所欲言、分享您的想法和见解。

推荐:Linux运维老纪的首页,持续学习,不断总结,共同进步,活到老学到老
导航剑指大厂系列:全面总结运维核心技术:系统基础、数据库、网路技术、系统安全、自动化运维、容器技术、监控工具、脚本编程、云服务等。
常用运维工具系列:常用的运维开发工具, zabbix、nagios、docker、k8s、puppet、ansible等
数据库系列:详细总结了常用数据库 mysql、Redis、MongoDB、oracle 技术点,以及工作中遇到的 mysql 问题等
懒人运维系列:总结好用的命令,解放双手不香吗?能用一个命令完成绝不用两个操作
数据结构与算法系列:总结数据结构和算法,不同类型针对性训练,提升编程思维,剑指大厂
非常期待和您一起在这个小小的网络世界里共同探索、学习和成长。💝💝💝 ✨✨ 欢迎订阅本专栏 ✨✨

kubernetes存储之GlusterFS

1、glusterfs概述

1.1、glusterfs简介

glusterfs是一个开源的、可扩展的分布式文件系统，集成来自多台服务器上的磁盘存储资源到单一全局命名空间，以提供共享文件存储。设计用于大型、分布式的应用，能够处理对大量数据的访问。‌ 它运行在廉价的普通硬件上，提供了高可用性和容错功能，确保了系统的稳定性和数据的可靠性。GFS由存储服务器、客户端以及NFS/Samba存储网关组成，没有元数据服务器组件，这有助于提升整个系统的性能、可靠性和稳定性。

1.2、glusterfs特点

可以扩展到几PB容量
支持处理数千个客户端
兼容POSIX接口
使用通用硬件，普通服务器即可构建
能够使用支持扩展属性的文件系统，例如ext4，XFS
支持工业标准的协议，例如NFS，SMB
提供很多高级功能，例如副本，配额，跨地域复制，快照以及bitrot检测
支持根据不同工作负载进行调优

1.3、glusterfs卷的模式

glusterfs中的volume的模式有很多中，包括以下几种：

分布卷（默认模式）：即DHT, 也叫分布卷: 将文件以hash算法随机分布到一台服务器节点中存储。
复制模式：即AFR, 创建volume 时带 replica x 数量: 将文件复制到 replica x 个节点中。
条带模式：即Striped, 创建volume 时带 stripe x 数量：将文件切割成数据块，分别存储到 stripe x 个节点中 ( 类似raid 0 )。
分布式条带模式：最少需要4台服务器才能创建。创建volume 时 stripe 2 server = 4 个节点：是DHT 与 Striped 的组合型。
分布式复制模式：最少需要4台服务器才能创建。创建volume 时 replica 2 server = 4 个节点：是DHT 与 AFR 的组合型。
条带复制卷模式：最少需要4台服务器才能创建。创建volume 时 stripe 2 replica 2 server = 4 个节点：是 Striped 与 AFR 的组合型。
三种模式混合：至少需要8台服务器才能创建。 stripe 2 replica 2 , 每4个节点组成一个组。

2、heketi概述

heketi是一个提供RESTful API管理gfs卷的框架，能够在kubernetes、openshift、openstack等云平台上实现动态的存储资源供应，支持gfs多集群管理，便于管理员对gfs进行操作，在kubernetes集群中，pod将存储的请求发送至heketi，然后heketi控制gfs集群创建对应的存储卷。
heketi动态在集群内选择bricks构建指定的volumes，以确保副本会分散到集群不同的故障域内。
heketi还支持任意数量的glusterfs集群，以保证接入的云服务器不局限于单个glusterfs集群。

3、部署heketi+glusterfs

环境：kubeadm安装的最新k8s 1.16.2版本，由1master+2node组成，网络插件选用的是flannel，默认kubeadm安装的k8s，会给master打上污点，本文为了实现gfs集群功能，先手动去掉了污点。
本文的glusterfs卷模式为复制卷模式。
另外，glusterfs在kubernetes集群中需要以特权运行，需要在kube-apiserver中添加–allow-privileged=true参数以开启此功能，默认此版本的kubeadm已开启。

[root@k8s-master-01 ~]# kubectl describe nodes k8s-master-01 |grep Taint
Taints:             node-role.kubernetes.io/master:NoSchedule  
[root@k8s-master-01 ~]# kubectl taint node k8s-master-01 node-role.kubernetes.io/master-
node/k8s-master-01 untainted
[root@k8s-master-01 ~]# kubectl describe nodes k8s-master-01 |grep Taint
Taints:             <none>

3.1、准备工作

为了保证pod能够正常使用gfs作为后端存储，需要每台运行pod的节点上提前安装gfs的客户端工具，其他存储方式也类似。

3.1.1、所有节点安装glusterfs客户端

$ yum install -y glusterfs glusterfs-fuse -y

3.1.2、节点打标签

需要安装gfs的kubernetes设置Label，因为gfs是通过kubernetes集群的DaemonSet方式安装的。
DaemonSet安装方式默认会在每个节点上都进行安装，除非安装前设置筛选要安装节点Label，带上此标签的节点才会安装。
安装脚本中设置DaemonSet中设置安装在贴有 storagenode=glusterfs的节点，所以这是事先将节点贴上对应Label。

[root@k8s-master-01 ~]# kubectl get nodes
NAME            STATUS   ROLES    AGE     VERSION
k8s-master-01   Ready    master   5d      v1.16.2
k8s-node-01     Ready    <none>   4d23h   v1.16.2
k8s-node-02     Ready    <none>   4d23h   v1.16.2
[root@k8s-master-01 ~]# kubectl label node k8s-master-01 storagenode=glusterfs
node/k8s-master-01 labeled
[root@k8s-master-01 ~]# kubectl label node k8s-node-01 storagenode=glusterfs  
node/k8s-node-01 labeled
[root@k8s-master-01 ~]# kubectl label node k8s-node-02 storagenode=glusterfs
node/k8s-node-02 labeled
[root@k8s-master-01 ~]# kubectl get nodes --show-labels 
NAME            STATUS   ROLES    AGE     VERSION   LABELS
k8s-master-01   Ready    master   5d      v1.16.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-master-01,kubernetes.io/os=linux,node-role.kubernetes.io/master=,storagenode=glusterfs
k8s-node-01     Ready    <none>   4d23h   v1.16.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-node-01,kubernetes.io/os=linux,storagenode=glusterfs
k8s-node-02     Ready    <none>   4d23h   v1.16.2   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-node-02,kubernetes.io/os=linux,storagenode=glusterfs

3.1.3、所有节点加载对应模块

$ modprobe dm_snapshot
$ modprobe dm_mirror
$ modprobe dm_thin_pool

查看是否加载

$ lsmod | grep dm_snapshot
$ lsmod | grep dm_mirror
$ lsmod | grep dm_thin_pool

3.2、创建glusterfs集群

采用容器化方式部署gfs集群，同样也可以使用传统方式部署，在生产环境中，gfs集群最好是独立于集群之外进行部署，之后只需要创建对应的endpoints即可。这里采用Daemonset方式部署，同时保证已经打上标签的节点上都运行一个gfs服务，并且均有提供存储的磁盘。

3.2.1、下载相关安装文件

[root@k8s-master-01 glusterfs]# pwd
/root/manifests/glusterfs
[root@k8s-master-01 glusterfs]# wget https://github.com/heketi/heketi/releases/download/v7.0.0/heketi-client-v7.0.0.linux.amd64.tar.gz
[root@k8s-master-01 glusterfs]# tar xf heketi-client-v7.0.0.linux.amd64.tar.gz
[root@k8s-master-01 glusterfs]# cd heketi-client/share/heketi/kubernetes/
[root@k8s-master-01 kubernetes]# pwd
/root/manifests/glusterfs/heketi-client/share/heketi/kubernetes

在本集群中，下面用到的daemonset控制器及后面用到的deployment控制器的api版本均变为了apps/v1，所以需要手动修改下载的json文件再进行部署，资源编排文件中需要指定selector声明。避免出现以下报错：

[root@k8s-master-01 kubernetes]# kubectl apply -f glusterfs-daemonset.json 
error: unable to recognize "glusterfs-daemonset.json": no matches for kind "DaemonSet" in version "extensions/v1beta1"

修改api版本

"apiVersion": "extensions/v1beta1"

为apps/v1

"apiVersion": "apps/v1",

指定selector声明

[root@k8s-master-01 kubernetes]# kubectl apply -f glusterfs-daemonset.json 
error: error validating "glusterfs-daemonset.json": error validating data: ValidationError(DaemonSet.spec): missing required field "selector" in io.k8s.api.apps.v1.DaemonSetSpec; if you choose to ignore these errors, turn validation off with --validate=false

对应后面内容的selector，用matchlabel相关联

"spec": {
    "selector": {
        "matchLabels": {
            "glusterfs-node": "daemonset"
        }
    },

3.2.2、创建集群

[root@k8s-master-01 kubernetes]# kubectl apply -f glusterfs-daemonset.json 
daemonset.apps/glusterfs created

注意：

这里使用的是默认的挂载方式，可使用其他磁盘作为gfs的工作目录

此处创建的namespace为default，可手动指定为其他namespace

3.2.3、查看gfs pods

[root@k8s-master-01 kubernetes]# kubectl get pods 
NAME              READY   STATUS    RESTARTS   AGE
glusterfs-9tttf   1/1     Running   0          1m10s
glusterfs-gnrnr   1/1     Running   0          1m10s
glusterfs-v92j5   1/1     Running   0          1m10s

3.3、创建heketi服务

3.3.1、创建heketi的service account对象

[root@k8s-master-01 kubernetes]# cat heketi-service-account.json 
{
  "apiVersion": "v1",
  "kind": "ServiceAccount",
  "metadata": {
    "name": "heketi-service-account"
  }
}
[root@k8s-master-01 kubernetes]# kubectl apply -f heketi-service-account.json 
serviceaccount/heketi-service-account created
[root@k8s-master-01 kubernetes]# kubectl get sa
NAME                     SECRETS   AGE
default                  1         71m
heketi-service-account   1         5s

3.3.2、创建heketi对应的权限和secret

[root@k8s-master-01 kubernetes]# kubectl create clusterrolebinding heketi-gluster-admin --clusterrole=edit --serviceaccount=dafault:heketi-service-account
clusterrolebinding.rbac.authorization.k8s.io/heketi-gluster-admin created
[root@k8s-master-01 kubernetes]# kubectl create secret generic heketi-config-secret --from-file=./heketi.json 
secret/heketi-config-secret created

3.3.3、初始化部署heketi

同样的，需要修改api版本以及增加selector声明部分。

[root@k8s-master-01 kubernetes]# vim heketi-bootstrap.json 
...
      "kind": "Deployment",
      "apiVersion": "apps/v1"
...
      "spec": {
        "selector": {
          "matchLabels": {
            "name": "deploy-heketi" 
          }
        },
...
[root@k8s-master-01 kubernetes]# kubectl create -f heketi-bootstrap.json 
service/deploy-heketi created
deployment.apps/deploy-heketi created
[root@k8s-master-01 kubernetes]# vim heketi-deployment.json
...
      "kind": "Deployment",
      "apiVersion": "apps/v1",
...
      "spec": {
        "selector": {
          "matchLabels": {
            "name": "heketi"
          }
        },
        "replicas": 1,
...
[root@k8s-master-01 kubernetes]# kubectl apply -f heketi-deployment.json
secret/heketi-db-backup created
service/heketi created
deployment.apps/heketi created
[root@k8s-master-01 kubernetes]# kubectl get pods 
NAME                             READY   STATUS    RESTARTS   AGE
deploy-heketi-6c687b4b84-p7mcr   1/1     Running   0          72s
heketi-68795ccd8-9726s           0/1     ContainerCreating   0          50s
glusterfs-9tttf                  1/1     Running   0          48m
glusterfs-gnrnr                  1/1     Running   0          48m
glusterfs-v92j5                  1/1     Running   0          48m

3.4、创建gfs集群

3.4.1、复制二进制文件

复制heketi-cli到/usr/local/bin目录下

[root@k8s-master-01 heketi-client]# pwd
/root/manifests/glusterfs/heketi-client
[root@k8s-master-01 heketi-client]# cp bin/heketi-cli /usr/local/bin/
[root@k8s-master-01 heketi-client]# heketi-cli -v
heketi-cli v7.0.0

3.4.2、配置topology-sample

修改topology-sample，manage为gfs管理服务的节点Node主机名，storage为节点的ip地址，device为节点上的裸设备，也就是用于提供存储的磁盘最好使用裸设备，不进行分区。
因此，需要预先在每个gfs的节点上准备好新的磁盘，这里分别在三个节点都新添加了一块/dev/sdb磁盘设备，大小均为10G。

[root@k8s-master-01 ~]# lsblk
NAME            MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sda               8:0    0   50G  0 disk 
├─sda1            8:1    0    2G  0 part /boot
└─sda2            8:2    0   48G  0 part 
  ├─centos-root 253:0    0   44G  0 lvm  /
  └─centos-swap 253:1    0    4G  0 lvm  
sdb               8:16   0   10G  0 disk 
sr0              11:0    1 1024M  0 rom
[root@k8s-node-01 ~]# lsblk
NAME            MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sda               8:0    0   50G  0 disk 
├─sda1            8:1    0    2G  0 part /boot
└─sda2            8:2    0   48G  0 part 
  ├─centos-root 253:0    0   44G  0 lvm  /
  └─centos-swap 253:1    0    4G  0 lvm  
sdb               8:16   0   10G  0 disk 
sr0              11:0    1 1024M  0 rom
[root@k8s-node-02 ~]# lsblk
NAME            MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sda               8:0    0   50G  0 disk 
├─sda1            8:1    0    2G  0 part /boot
└─sda2            8:2    0   48G  0 part 
  ├─centos-root 253:0    0   44G  0 lvm  /
  └─centos-swap 253:1    0    4G  0 lvm  
sdb               8:16   0   10G  0 disk 
sr0              11:0    1 1024M  0 rom

配置topology-sample

{
    "clusters": [
        {
            "nodes": [
                {
                    "node": {
                        "hostnames": {
                            "manage": [
                                "k8s-master-01"
                            ],
                            "storage": [
                                "192.168.2.10"
                            ]
                        },
                        "zone": 1
                    },
                    "devices": [
                        {
                            "name": "/dev/sdb",
                            "destroydata": false
                        }
                    ]
                },
                {
                    "node": {
                        "hostnames": {
                            "manage": [
                                "k8s-node-01"
                            ],
                            "storage": [
                                "192.168.2.11"
                            ]
                        },
                        "zone": 1
                    },
                    "devices": [
                        {
                            "name": "/dev/sdb",
                            "destroydata": false
                        }
                    ]
                },
                {
                    "node": {
                        "hostnames": {
                            "manage": [
                                "k8s-node-02"
                            ],
                            "storage": [
                                "192.168.2.12"
                            ]
                        },
                        "zone": 1
                    },
                    "devices": [
                        {
                            "name": "/dev/sdb",
                            "destroydata": false
                        }
                    ]
                }
            ]
        }
    ]
}

3.4.3、获取当前heketi的ClusterIP

查看当前heketi的ClusterIP，并通过环境变量声明

[root@k8s-master-01 kubernetes]# kubectl get svc|grep heketi
deploy-heketi   ClusterIP   10.1.241.99   <none>        8080/TCP   3m18s
[root@k8s-master-01 kubernetes]# curl http://10.1.241.99:8080/hello
Hello from Heketi
[root@k8s-master-01 kubernetes]# export HEKETI_CLI_SERVER=http://10.1.241.99:8080
[root@k8s-master-01 kubernetes]# echo $HEKETI_CLI_SERVER
http://10.1.185.215:8080

3.4.4、使用heketi创建gfs集群

执行如下命令创建gfs集群会提示Invalid JWT token: Token missing iss claim

[root@k8s-master-01 kubernetes]# heketi-cli topology load --json=topology-sample.json 
Error: Unable to get topology information: Invalid JWT token: Token missing iss claim

这是因为新版本的heketi在创建gfs集群时需要带上参数，声明用户名及密码，相应值在heketi.json文件中配置，即：

[root@k8s-master-01 kubernetes]# heketi-cli -s $HEKETI_CLI_SERVER --user admin --secret 'My Secret' topology load --json=topology-sample.json
Creating cluster ... ID: 1c5ffbd86847e5fc1562ef70c033292e
        Allowing file volumes on cluster.
        Allowing block volumes on cluster.
        Creating node k8s-master-01 ... ID: b6100a5af9b47d8c1f19be0b2b4d8276
                Adding device /dev/sdb ... OK
        Creating node k8s-node-01 ... ID: 04740cac8d42f56e354c94bdbb7b8e34
                Adding device /dev/sdb ... OK
        Creating node k8s-node-02 ... ID: 1b33ad0dba20eaf23b5e3a4845e7cdb4
                Adding device /dev/sdb ... OK

执行了heketi-cli topology load之后，Heketi在服务器做的大致操作如下：

进入任意glusterfs Pod内，执行gluster peer status 发现都已把对端加入到了可信存储池(TSP)中。
在运行了gluster Pod的节点上，自动创建了一个VG，此VG正是由topology-sample.json 文件中的磁盘裸设备创建而来。
一块磁盘设备创建出一个VG，以后创建的PVC，即从此VG里划分的LV。
heketi-cli topology info 查看拓扑结构，显示出每个磁盘设备的ID，对应VG的ID，总空间、已用空间、空余空间等信息。
通过部分日志查看

[root@k8s-master-01 manifests]# kubectl logs -f deploy-heketi-6c687b4b84-l5b6j
...
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [pvs -o pv_name,pv_uuid,vg_name --reportformat=json /dev/sdb] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout [  {
      "report": [
          {
              "pv": [
                  {"pv_name":"/dev/sdb", "pv_uuid":"1UkSIV-RYt1-QBNw-KyAR-Drm5-T9NG-UmO313", "vg_name":"vg_398329cc70361dfd4baa011d811de94a"}
              ]
          }
      ]
  }
]: Stderr [  WARNING: Device /dev/sdb not initialized in udev database even after waiting 10000000 microseconds.
  WARNING: Device /dev/centos/root not initialized in udev database even after waiting 10000000 microseconds.
  WARNING: Device /dev/sda1 not initialized in udev database even after waiting 10000000 microseconds.
  WARNING: Device /dev/centos/swap not initialized in udev database even after waiting 10000000 microseconds.
  WARNING: Device /dev/sda2 not initialized in udev database even after waiting 10000000 microseconds.
  WARNING: Device /dev/sdb not initialized in udev database even after waiting 10000000 microseconds.
]
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [udevadm info --query=symlink --name=/dev/sdb] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [udevadm info --query=symlink --name=/dev/sdb] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout [
]: Stderr []
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [vgdisplay -c vg_398329cc70361dfd4baa011d811de94a] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[negroni] 2019-10-23T02:17:44Z | 200 |   93.868µs | 10.1.241.99:8080 | GET /queue/3d0b6edb0faa67e8efd752397f314a6f
[kubeexec] DEBUG 2019/10/23 02:17:44 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [vgdisplay -c vg_398329cc70361dfd4baa011d811de94a] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout [  vg_398329cc70361dfd4baa011d811de94a:r/w:772:-1:0:0:0:-1:0:1:1:10350592:4096:2527:0:2527:YCPG9X-b270-1jf2-VwKX-ycpZ-OI9u-7ZidOc
]: Stderr []
[cmdexec] DEBUG 2019/10/23 02:17:44 heketi/executors/cmdexec/device.go:273:cmdexec.(*CmdExecutor).getVgSizeFromNode: /dev/sdb in k8s-node-01 has TotalSize:10350592, FreeSize:10350592, UsedSize:0
[heketi] INFO 2019/10/23 02:17:44 Added device /dev/sdb
[asynchttp] INFO 2019/10/23 02:17:44 Completed job 3d0b6edb0faa67e8efd752397f314a6f in 3m2.694238221s
[negroni] 2019-10-23T02:17:45Z | 204 |   105.23µs | 10.1.241.99:8080 | GET /queue/3d0b6edb0faa67e8efd752397f314a6f
[cmdexec] INFO 2019/10/23 02:17:45 Check Glusterd service status in node k8s-node-01
[kubeexec] DEBUG 2019/10/23 02:17:45 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [systemctl status glusterd] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:45 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 02:17:45 heketi/pkg/remoteexec/log/commandlog.go:41:log.(*CommandLogger).Success: Ran command [systemctl status glusterd] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout filtered, Stderr filtered
[heketi] INFO 2019/10/23 02:17:45 Adding node k8s-node-02
[negroni] 2019-10-23T02:17:45Z | 202 |   146.998544ms | 10.1.241.99:8080 | POST /nodes
[asynchttp] INFO 2019/10/23 02:17:45 Started job 8da70b6fd6fec1d61c4ba1cd0fe27fe5
[cmdexec] INFO 2019/10/23 02:17:45 Probing: k8s-node-01 -> 192.168.2.12
[negroni] 2019-10-23T02:17:45Z | 200 |   74.577µs | 10.1.241.99:8080 | GET /queue/8da70b6fd6fec1d61c4ba1cd0fe27fe5
[kubeexec] DEBUG 2019/10/23 02:17:45 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [gluster --mode=script --timeout=600 peer probe 192.168.2.12] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:45 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[negroni] 2019-10-23T02:17:46Z | 200 |   79.893µs | 10.1.241.99:8080 | GET /queue/8da70b6fd6fec1d61c4ba1cd0fe27fe5
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [gluster --mode=script --timeout=600 peer probe 192.168.2.12] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout [peer probe: success. 
]: Stderr []
[cmdexec] INFO 2019/10/23 02:17:46 Setting snapshot limit
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [gluster --mode=script --timeout=600 snapshot config snap-max-hard-limit 14] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [gluster --mode=script --timeout=600 snapshot config snap-max-hard-limit 14] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout [snapshot config: snap-max-hard-limit for System set successfully
]: Stderr []
[heketi] INFO 2019/10/23 02:17:46 Added node 1b33ad0dba20eaf23b5e3a4845e7cdb4
[asynchttp] INFO 2019/10/23 02:17:46 Completed job 8da70b6fd6fec1d61c4ba1cd0fe27fe5 in 488.404011ms
[negroni] 2019-10-23T02:17:46Z | 303 |   80.712µs | 10.1.241.99:8080 | GET /queue/8da70b6fd6fec1d61c4ba1cd0fe27fe5
[negroni] 2019-10-23T02:17:46Z | 200 |   242.595µs | 10.1.241.99:8080 | GET /nodes/1b33ad0dba20eaf23b5e3a4845e7cdb4
[heketi] INFO 2019/10/23 02:17:46 Adding device /dev/sdb to node 1b33ad0dba20eaf23b5e3a4845e7cdb4
[negroni] 2019-10-23T02:17:46Z | 202 |   696.018µs | 10.1.241.99:8080 | POST /devices
[asynchttp] INFO 2019/10/23 02:17:46 Started job 21af2069b74762a5521a46e2b52e7d6a
[negroni] 2019-10-23T02:17:46Z | 200 |   82.354µs | 10.1.241.99:8080 | GET /queue/21af2069b74762a5521a46e2b52e7d6a
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [pvcreate -qq --metadatasize=128M --dataalignment=256K '/dev/sdb'] on [pod:glusterfs-l2lsv c:glusterfs ns:default (from host:k8s-node-02 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 02:17:46 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
...

3.4.5、持久化heketi配置

上面创建的heketi没有配置持久化的卷，如果heketi的pod重启，可能会丢失之前的配置信息，所以现在创建heketi持久化的卷来对heketi数据进行持久化，该持久化方式利用gfs提供的动态存储，也可以采用其他方式进行持久化。
在所有节点安装device-mapper*

yum install -y device-mapper*

将配置信息保存为文件，并创建持久化相关信息

[root@k8s-master-01 kubernetes]# heketi-cli -s $HEKETI_CLI_SERVER --user admin --secret 'My Secret' setup-openshift-heketi-storage Saving heketi-storage.json
Saving heketi-storage.json
[root@k8s-master-01 kubernetes]# kubectl apply -f heketi-storage.json 
secret/heketi-storage-secret created
endpoints/heketi-storage-endpoints created
service/heketi-storage-endpoints created
job.batch/heketi-storage-copy-job created

删除中间产物

[root@k8s-master-01 kubernetes]# kubectl delete all,svc,jobs,deployment,secret --selector="deploy-heketi"
pod "deploy-heketi-6c687b4b84-l5b6j" deleted
service "deploy-heketi" deleted
deployment.apps "deploy-heketi" deleted
replicaset.apps "deploy-heketi-6c687b4b84" deleted
job.batch "heketi-storage-copy-job" deleted
secret "heketi-storage-secret" deleted

创建持久化的heketi

[root@k8s-master-01 kubernetes]# kubectl apply -f heketi-deployment.json 
secret/heketi-db-backup created
service/heketi created
deployment.apps/heketi created
[root@k8s-master-01 kubernetes]# kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
glusterfs-cqw5d          1/1     Running   0          41m
glusterfs-l2lsv          1/1     Running   0          41m
glusterfs-lrdz7          1/1     Running   0          41m
heketi-68795ccd8-m8x55   1/1     Running   0          32s

查看持久化后heketi的svc，并重新声明环境变量

[root@k8s-master-01 kubernetes]# kubectl get svc
NAME                       TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)    AGE
heketi                     ClusterIP   10.1.45.61   <none>        8080/TCP   2m9s
heketi-storage-endpoints   ClusterIP   10.1.26.73   <none>        1/TCP      4m58s
kubernetes                 ClusterIP   10.1.0.1     <none>        443/TCP    14h
[root@k8s-master-01 kubernetes]# export HEKETI_CLI_SERVER=http://10.1.45.61:8080
[root@k8s-master-01 kubernetes]# curl http://10.1.45.61:8080/hello
Hello from Heketi

查看gfs集群信息，更多操作参照官方文档说明

[root@k8s-master-01 kubernetes]# heketi-cli -s $HEKETI_CLI_SERVER --user admin --secret 'My Secret' topology info

Cluster Id: 1c5ffbd86847e5fc1562ef70c033292e

    File:  true
    Block: true

    Volumes:

        Name: heketidbstorage
        Size: 2
        Id: b25f4b627cf66279bfe19e8a01e9e85d
        Cluster Id: 1c5ffbd86847e5fc1562ef70c033292e
        Mount: 192.168.2.11:heketidbstorage
        Mount Options: backup-volfile-servers=192.168.2.12,192.168.2.10
        Durability Type: replicate
        Replica: 3
        Snapshot: Disabled

                Bricks:
                        Id: 3ab6c19b8fe0112575ba04d58573a404
                        Path: /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_3ab6c19b8fe0112575ba04d58573a404/brick
                        Size (GiB): 2
                        Node: b6100a5af9b47d8c1f19be0b2b4d8276
                        Device: 703e3662cbd8ffb24a6401bb3c3c41fa

                        Id: d1fa386f2ec9954f4517431163f67dea
                        Path: /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_d1fa386f2ec9954f4517431163f67dea/brick
                        Size (GiB): 2
                        Node: 04740cac8d42f56e354c94bdbb7b8e34
                        Device: 398329cc70361dfd4baa011d811de94a

                        Id: d2b0ae26fa3f0eafba407b637ca0d06b
                        Path: /var/lib/heketi/mounts/vg_7c791bbb90f710123ba431a7cdde8d0b/brick_d2b0ae26fa3f0eafba407b637ca0d06b/brick
                        Size (GiB): 2
                        Node: 1b33ad0dba20eaf23b5e3a4845e7cdb4
                        Device: 7c791bbb90f710123ba431a7cdde8d0b



    Nodes:

        Node Id: 04740cac8d42f56e354c94bdbb7b8e34
        State: online
        Cluster Id: 1c5ffbd86847e5fc1562ef70c033292e
        Zone: 1
        Management Hostnames: k8s-node-01
        Storage Hostnames: 192.168.2.11
        Devices:
                Id:398329cc70361dfd4baa011d811de94a   Name:/dev/sdb            State:online    Size (GiB):9       Used (GiB):2       Free (GiB):7       
                        Bricks:
                                Id:d1fa386f2ec9954f4517431163f67dea   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_d1fa386f2ec9954f4517431163f67dea/brick

        Node Id: 1b33ad0dba20eaf23b5e3a4845e7cdb4
        State: online
        Cluster Id: 1c5ffbd86847e5fc1562ef70c033292e
        Zone: 1
        Management Hostnames: k8s-node-02
        Storage Hostnames: 192.168.2.12
        Devices:
                Id:7c791bbb90f710123ba431a7cdde8d0b   Name:/dev/sdb            State:online    Size (GiB):9       Used (GiB):2       Free (GiB):7       
                        Bricks:
                                Id:d2b0ae26fa3f0eafba407b637ca0d06b   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_7c791bbb90f710123ba431a7cdde8d0b/brick_d2b0ae26fa3f0eafba407b637ca0d06b/brick

        Node Id: b6100a5af9b47d8c1f19be0b2b4d8276
        State: online
        Cluster Id: 1c5ffbd86847e5fc1562ef70c033292e
        Zone: 1
        Management Hostnames: k8s-master-01
        Storage Hostnames: 192.168.2.10
        Devices:
                Id:703e3662cbd8ffb24a6401bb3c3c41fa   Name:/dev/sdb            State:online    Size (GiB):9       Used (GiB):2       Free (GiB):7       
                        Bricks:
                                Id:3ab6c19b8fe0112575ba04d58573a404   Size (GiB):2       Path: /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_3ab6c19b8fe0112575ba04d58573a404/brick

4、创建storageclass

[root@k8s-master-01 kubernetes]# vim storageclass-gfs-heketi.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: gluster-heketi
provisioner: kubernetes.io/glusterfs
reclaimPolicy: Retain
parameters:
  resturl: "http://10.1.45.61:8080"
  restauthenabled: "true"
  restuser: "admin"
  restuserkey: "My Secret"
  gidMin: "40000"
  gidMax: "50000"
  volumetype: "replicate:3"
allowVolumeExpansion: true
[root@k8s-master-01 kubernetes]# kubectl apply -f storageclass-gfs-heketi.yaml
storageclass.storage.k8s.io/gluster-heketi created

参数说明：

reclaimPolicy：Retain 回收策略，默认是Delete，删除pvc后pv及后端创建的volume、brick(lvm)不会被删除。
gidMin和gidMax，能够使用的最小和最大gid
volumetype：卷类型及个数，这里使用的是复制卷，个数必须大于1

5、测试通过gfs提供动态存储

创建一个pod使用动态pv，在StorageClassName指定之前创建的StorageClass的name，即gluster-heketi：

[root@k8s-master-01 kubernetes]# vim pod-use-pvc.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: pod-use-pvc
spec:
  containers:
  - name: pod-use-pvc
    image: busybox
    command:
      - sleep
      - "3600"
    volumeMounts:
    - name: gluster-volume
      mountPath: "/pv-data"
      readOnly: false
  volumes:
  - name: gluster-volume
    persistentVolumeClaim:
      claimName: pvc-gluster-heketi

---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: pvc-gluster-heketi
spec:
  accessModes: [ "ReadWriteOnce" ]
  storageClassName: "gluster-heketi"
  resources:
    requests:
      storage: 1Gi

创建pod并查看创建的pv和pvc

[root@k8s-master-01 kubernetes]# kubectl apply -f pod-use-pvc.yaml 
pod/pod-use-pvc created
persistentvolumeclaim/pvc-gluster-heketi created
[root@k8s-master-01 kubernetes]# kubectl get pv,pvc
NAME                                                        CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                        STORAGECLASS     REASON   AGE
persistentvolume/pvc-0fb9b246-4da4-491c-b6a2-4f38489ab11c   1Gi        RWO            Retain           Bound    default/pvc-gluster-heketi   gluster-heketi            57s

NAME                                       STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS     AGE
persistentvolumeclaim/pvc-gluster-heketi   Bound    pvc-0fb9b246-4da4-491c-b6a2-4f38489ab11c   1Gi        RWO            gluster-heketi   62s

6、分析k8s通过heketi创建pv及pvc的过程

通过pvc及向storageclass申请创建对应的pv，具体可通过查看创建的heketi pod的日志
首先发现heketi接收到请求之后运行了一个job任务，创建了三个bricks，在三个gfs节点中创建对应的目录：

[heketi] INFO 2019/10/23 03:08:36 Allocating brick set #0
[negroni] 2019-10-23T03:08:36Z | 202 |   56.193603ms | 10.1.45.61:8080 | POST /volumes
[asynchttp] INFO 2019/10/23 03:08:36 Started job 3ec932315085609bc54ead6e3f6851e8
[heketi] INFO 2019/10/23 03:08:36 Started async operation: Create Volume
[heketi] INFO 2019/10/23 03:08:36 Trying Create Volume (attempt #1/5)
[heketi] INFO 2019/10/23 03:08:36 Creating brick 289fe032c1f4f9f211480e24c5d74a44
[heketi] INFO 2019/10/23 03:08:36 Creating brick a3172661ba1b849d67b500c93c3dd652
[heketi] INFO 2019/10/23 03:08:36 Creating brick 917e27a9dbc5395ebf08dff8d3401b43
[negroni] 2019-10-23T03:08:36Z | 200 |   72.083µs | 10.1.45.61:8080 | GET /queue/3ec932315085609bc54ead6e3f6851e8
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [mkdir -p /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_a3172661ba1b849d67b500c93c3dd652] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [mkdir -p /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_289fe032c1f4f9f211480e24c5d74a44] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 1
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [mkdir -p /var/lib/heketi/mounts/vg_7c791bbb90f710123ba431a7cdde8d0b/brick_917e27a9dbc5395ebf08dff8d3401b43] on [pod:glusterfs-l2lsv c:glusterfs ns:default (from host:k8s-node-02 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:08:36 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 2

创建lv，添加自动挂载

[kubeexec] DEBUG 2019/10/23 03:08:37 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 2
[kubeexec] DEBUG 2019/10/23 03:08:37 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [mkfs.xfs -i size=512 -n size=8192 /dev/mapper/vg_703e3662cbd8ffb24a6401bb3c3c41fa-brick_a3172661ba1b849d67b500c93c3dd652] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]: Stdout [meta-data=/dev/mapper/vg_703e3662cbd8ffb24a6401bb3c3c41fa-brick_a3172661ba1b849d67b500c93c3dd652 isize=512    agcount=8, agsize=32768 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=0, sparse=0
data     =                       bsize=4096   blocks=262144, imaxpct=25
         =                       sunit=64     swidth=64 blks
naming   =version 2              bsize=8192   ascii-ci=0 ftype=1
log      =internal log           bsize=4096   blocks=2560, version=2
         =                       sectsz=512   sunit=64 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
]: Stderr []
[kubeexec] DEBUG 2019/10/23 03:08:37 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [awk "BEGIN {print \"/dev/mapper/vg_703e3662cbd8ffb24a6401bb3c3c41fa-brick_a3172661ba1b849d67b500c93c3dd652 /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_a3172661ba1b849d67b500c93c3dd652 xfs rw,inode64,noatime,nouuid 1 2\" >> \"/var/lib/heketi/fstab\"}"] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]

创建brick，设置权限

[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [chmod 2775 /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_a3172661ba1b849d67b500c93c3dd652/brick] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 2
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [chown :40000 /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_289fe032c1f4f9f211480e24c5d74a44/brick] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout []: Stderr []
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [chmod 2775 /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_289fe032c1f4f9f211480e24c5d74a44/brick] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 2
[negroni] 2019-10-23T03:08:38Z | 200 |   83.159µs | 10.1.45.61:8080 | GET /queue/3ec932315085609bc54ead6e3f6851e8
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [chmod 2775 /var/lib/heketi/mounts/vg_7c791bbb90f710123ba431a7cdde8d0b/brick_917e27a9dbc5395ebf08dff8d3401b43/brick] on [pod:glusterfs-l2lsv c:glusterfs ns:default (from host:k8s-node-02 selector:glusterfs-node)]: Stdout []: Stderr []
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [chmod 2775 /var/lib/heketi/mounts/vg_703e3662cbd8ffb24a6401bb3c3c41fa/brick_a3172661ba1b849d67b500c93c3dd652/brick] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]: Stdout []: Stderr []
[kubeexec] DEBUG 2019/10/23 03:08:38 heketi/pkg/remoteexec/log/commandlog.go:46:log.(*CommandLogger).Success: Ran command [chmod 2775 /var/lib/heketi/mounts/vg_398329cc70361dfd4baa011d811de94a/brick_289fe032c1f4f9f211480e24c5d74a44/brick] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout []: Stderr []
[cmdexec] INFO 2019/10/23 03:08:38 Creating volume vol_08e8447256de2598952dcb240e615d0f replica 3

创建对应的volume

[asynchttp] INFO 2019/10/23 03:08:41 Completed job 3ec932315085609bc54ead6e3f6851e8 in 5.007631648s
[negroni] 2019-10-23T03:08:41Z | 303 |   78.335µs | 10.1.45.61:8080 | GET /queue/3ec932315085609bc54ead6e3f6851e8
[negroni] 2019-10-23T03:08:41Z | 200 |   5.751689ms | 10.1.45.61:8080 | GET /volumes/08e8447256de2598952dcb240e615d0f
[negroni] 2019-10-23T03:08:41Z | 200 |   139.05µs | 10.1.45.61:8080 | GET /clusters/1c5ffbd86847e5fc1562ef70c033292e
[negroni] 2019-10-23T03:08:41Z | 200 |   660.249µs | 10.1.45.61:8080 | GET /nodes/04740cac8d42f56e354c94bdbb7b8e34
[negroni] 2019-10-23T03:08:41Z | 200 |   270.334µs | 10.1.45.61:8080 | GET /nodes/1b33ad0dba20eaf23b5e3a4845e7cdb4
[negroni] 2019-10-23T03:08:41Z | 200 |   345.528µs | 10.1.45.61:8080 | GET /nodes/b6100a5af9b47d8c1f19be0b2b4d8276
[heketi] INFO 2019/10/23 03:09:39 Starting Node Health Status refresh
[cmdexec] INFO 2019/10/23 03:09:39 Check Glusterd service status in node k8s-node-01
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [systemctl status glusterd] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:41:log.(*CommandLogger).Success: Ran command [systemctl status glusterd] on [pod:glusterfs-lrdz7 c:glusterfs ns:default (from host:k8s-node-01 selector:glusterfs-node)]: Stdout filtered, Stderr filtered
[heketi] INFO 2019/10/23 03:09:39 Periodic health check status: node 04740cac8d42f56e354c94bdbb7b8e34 up=true
[cmdexec] INFO 2019/10/23 03:09:39 Check Glusterd service status in node k8s-node-02
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [systemctl status glusterd] on [pod:glusterfs-l2lsv c:glusterfs ns:default (from host:k8s-node-02 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:41:log.(*CommandLogger).Success: Ran command [systemctl status glusterd] on [pod:glusterfs-l2lsv c:glusterfs ns:default (from host:k8s-node-02 selector:glusterfs-node)]: Stdout filtered, Stderr filtered
[heketi] INFO 2019/10/23 03:09:39 Periodic health check status: node 1b33ad0dba20eaf23b5e3a4845e7cdb4 up=true
[cmdexec] INFO 2019/10/23 03:09:39 Check Glusterd service status in node k8s-master-01
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:34:log.(*CommandLogger).Before: Will run command [systemctl status glusterd] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/kube/exec.go:72:kube.ExecCommands: Current kube connection count: 0
[kubeexec] DEBUG 2019/10/23 03:09:39 heketi/pkg/remoteexec/log/commandlog.go:41:log.(*CommandLogger).Success: Ran command [systemctl status glusterd] on [pod:glusterfs-cqw5d c:glusterfs ns:default (from host:k8s-master-01 selector:glusterfs-node)]: Stdout filtered, Stderr filtered
[heketi] INFO 2019/10/23 03:09:39 Periodic health check status: node b6100a5af9b47d8c1f19be0b2b4d8276 up=true
[heketi] INFO 2019/10/23 03:09:39 Cleaned 0 nodes from health cache

7、测试数据

测试使用该pv的pod之间能否共享数据，手动进入到pod并创建文件

[root@k8s-master-01 kubernetes]# kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
glusterfs-cqw5d          1/1     Running   0          90m
glusterfs-l2lsv          1/1     Running   0          90m
glusterfs-lrdz7          1/1     Running   0          90m
heketi-68795ccd8-m8x55   1/1     Running   0          49m
pod-use-pvc              1/1     Running   0          20m
[root@k8s-master-01 kubernetes]# kubectl exec -it pod-use-pvc /bin/sh
/ # cd /pv-data/
/pv-data # echo "hello world">a.txt
/pv-data # cat a.txt 
hello world

查看创建的卷

[root@k8s-master-01 kubernetes]# heketi-cli -s $HEKETI_CLI_SERVER --user admin --secret 'My Secret' volume list
Id:08e8447256de2598952dcb240e615d0f    Cluster:1c5ffbd86847e5fc1562ef70c033292e    Name:vol_08e8447256de2598952dcb240e615d0f
Id:b25f4b627cf66279bfe19e8a01e9e85d    Cluster:1c5ffbd86847e5fc1562ef70c033292e    Name:heketidbstorage

将设备挂载查看卷中的数据，vol_08e8447256de2598952dcb240e615d0f为卷名称

[root@k8s-master-01 kubernetes]# mount -t glusterfs 192.168.2.10:vol_08e8447256de2598952dcb240e615d0f /mnt
[root@k8s-master-01 kubernetes]# ll /mnt/
total 1
-rw-r--r-- 1 root 40000 12 Oct 23 11:29 a.txt
[root@k8s-master-01 kubernetes]# cat /mnt/a.txt 
hello world

8、测试deployment

测试通过deployment控制器部署能否正常使用storageclass，创建nginx的deployment

[root@k8s-master-01 kubernetes]# vim nginx-deployment-gluster.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-gfs
spec:
  selector:
    matchLabels:
      name: nginx
  replicas: 2
  template:
    metadata:
      labels:
        name: nginx
    spec:
      containers:
        - name: nginx
          image: nginx
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 80
          volumeMounts:
            - name: nginx-gfs-html
              mountPath: "/usr/share/nginx/html"
            - name: nginx-gfs-conf
              mountPath: "/etc/nginx/conf.d"
      volumes:
      - name: nginx-gfs-html
        persistentVolumeClaim:
          claimName: glusterfs-nginx-html
      - name: nginx-gfs-conf
        persistentVolumeClaim:
          claimName: glusterfs-nginx-conf

---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: glusterfs-nginx-html
spec:
  accessModes: [ "ReadWriteMany" ]
  storageClassName: "gluster-heketi"
  resources:
    requests:
      storage: 500Mi

---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: glusterfs-nginx-conf
spec:
  accessModes: [ "ReadWriteMany" ]
  storageClassName: "gluster-heketi"
  resources:
    requests:
      storage: 10Mi

查看相应资源

[root@k8s-master-01 kubernetes]# kubectl get pod,pv,pvc|grep nginx
pod/nginx-gfs-7d66cccf76-mkc76   1/1     Running   0          2m45s
pod/nginx-gfs-7d66cccf76-zc8n2   1/1     Running   0          2m45s
persistentvolume/pvc-87481e3a-9b7e-43aa-a0b9-4028ce0a1abb   1Gi        RWX            Retain           Bound    default/glusterfs-nginx-conf   gluster-heketi            2m34s
persistentvolume/pvc-f954a4ca-ea1c-458d-8490-a49a0a001ab5   1Gi        RWX            Retain           Bound    default/glusterfs-nginx-html   gluster-heketi            2m34s
persistentvolumeclaim/glusterfs-nginx-conf   Bound    pvc-87481e3a-9b7e-43aa-a0b9-4028ce0a1abb   1Gi        RWX            gluster-heketi   2m45s
persistentvolumeclaim/glusterfs-nginx-html   Bound    pvc-f954a4ca-ea1c-458d-8490-a49a0a001ab5   1Gi        RWX            gluster-heketi   2m45s

查看挂载情况

[root@k8s-master-01 kubernetes]# kubectl exec -it nginx-gfs-7d66cccf76-mkc76 -- df -Th
Filesystem                                        Type            Size  Used Avail Use% Mounted on
overlay                                           overlay          44G  3.2G   41G   8% /
tmpfs                                             tmpfs            64M     0   64M   0% /dev
tmpfs                                             tmpfs           2.0G     0  2.0G   0% /sys/fs/cgroup
/dev/mapper/centos-root                           xfs              44G  3.2G   41G   8% /etc/hosts
shm                                               tmpfs            64M     0   64M   0% /dev/shm
192.168.2.10:vol_adf6fc08c8828fdda27c8aa5ce99b50c fuse.glusterfs 1014M   43M  972M   5% /etc/nginx/conf.d
192.168.2.10:vol_454e14ae3184122ff9a14d77e02b10b9 fuse.glusterfs 1014M   43M  972M   5% /usr/share/nginx/html
tmpfs                                             tmpfs           2.0G   12K  2.0G   1% /run/secrets/kubernetes.io/serviceaccount
tmpfs                                             tmpfs           2.0G     0  2.0G   0% /proc/acpi
tmpfs                                             tmpfs           2.0G     0  2.0G   0% /proc/scsi
tmpfs                                             tmpfs           2.0G     0  2.0G   0% /sys/firmware

在宿主机挂载和创建文件

[root@k8s-master-01 kubernetes]# mount -t glusterfs 192.168.2.10:vol_454e14ae3184122ff9a14d77e02b10b9 /mnt/
[root@k8s-master-01 kubernetes]# cd /mnt/
[root@k8s-master-01 mnt]# echo "hello world">index.html
[root@k8s-master-01 mnt]# kubectl exec -it nginx-gfs-7d66cccf76-mkc76 -- cat /usr/share/nginx/html/index.html
hello world

扩容nginx副本，查看是否能正常挂载

[root@k8s-master-01 mnt]# kubectl scale deployment nginx-gfs --replicas=3
deployment.apps/nginx-gfs scaled
[root@k8s-master-01 mnt]# kubectl get pods
NAME                         READY   STATUS    RESTARTS   AGE
glusterfs-cqw5d              1/1     Running   0          129m
glusterfs-l2lsv              1/1     Running   0          129m
glusterfs-lrdz7              1/1     Running   0          129m
heketi-68795ccd8-m8x55       1/1     Running   0          88m
nginx-gfs-7d66cccf76-mkc76   1/1     Running   0          8m55s
nginx-gfs-7d66cccf76-qzqnv   1/1     Running   0          23s
nginx-gfs-7d66cccf76-zc8n2   1/1     Running   0          8m55s
[root@k8s-master-01 mnt]# kubectl exec -it nginx-gfs-7d66cccf76-qzqnv -- cat /usr/share/nginx/html/index.html     
hello world

至此，在k8s集群中部署heketi+glusterfs提供动态存储结束。