kubernetes存储之GlusterFS
目录
1、glusterfs概述
1.1、glusterfs简介
glusterfs是一个可扩展，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。
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。
参考来源：
https://github.com/heketi/heketi
https://github.com/gluster/gluster-kubernetes
https://www.cnblogs.com/jicki/p/5801712.html
kubernetes存储之GlusterFS的更多相关文章
- kubernetes 存储卷
kubernetes 存储卷 数据卷用于实现容器持久化数据,Kubernetes对于数据卷重新定义,提供了丰富强大的功能.在Kubernetes系统中,当Pod重建的时候,数据卷会丢失,Kube ...
- 附009.Kubernetes永久存储之GlusterFS独立部署
一 前期准备 1.1 基础知识 Heketi提供了一个RESTful管理界面,可以用来管理GlusterFS卷的生命周期.Heketi会动态在集群内选择bricks构建所需的volumes,从而确保数 ...
- 附010.Kubernetes永久存储之GlusterFS超融合部署
一 前期准备 1.1 基础知识 在Kubernetes中,使用GlusterFS文件系统,操作步骤通常是: 创建brick-->创建volume-->创建PV-->创建PVC--&g ...
- (九)Kubernetes 存储卷
Kubernetes存储卷概述 Pod本身具有生命周期,这就带了一系列的问题,第一,当一个容器损坏之后,kubelet会重启这个容器,但是文件会丢失-这个容器会是一个全新的状态:第二,当很多容器在同一 ...
- kubernetes资源使用glusterfs卷进行数据持久化
1.GlusterFS部署 安装GlusterFS集群的主要目的是为k8s集群提供分布式持久化存储. GlusterFS部署使用2台服务器,服务名称与IP如下: 1 db-storagea 10.1. ...
- 从零开始入门 K8s | Kubernetes 存储架构及插件使用
本文整理自<CNCF x Alibaba 云原生技术公开课>第 21 讲. 导读:容器存储是 Kubernetes 系统中提供数据持久化的基础组件,是实现有状态服务的重要保证.Kubern ...
- 第21 章 ： Kubernetes 存储架构及插件使用
Kubernetes 存储架构及插件使用 本文将主要分享以下三方面的内容: Kubernetes 存储体系架构: Flexvolume 介绍及使用: CSI 介绍及使用. Kubernetes 存储体 ...
- 详解Kubernetes存储体系
Volume.PV.PVC.StorageClass由来 先思考一个问题,为什么会引入Volume这样一个概念? " 答案很简单,为了实现数据持久化,数据的生命周期不随着容器的消亡而消亡. ...
- Kubernetes存储之Persistent Volumes简介
简介 管理存储和管理计算有着明显的不同.PersistentVolume子系统给用户和管理员提供了一套API,从而抽象出存储是如何提供和消耗的细节.在这里,我们介绍两种新的API资源:Persiste ...
随机推荐
- syncronizationcontext
FormDesign.cs 增加变量定义 public string testName { get; set; } TaskWrapper.cs string tmpTest = &qu ...
- 大数据开发工程师面试《一》Shopee虾皮技术面
一.项目问题 1 做了哪些项目2 使用什么技术3 哪个是你主导的项目,一共开发多少个接口,项目多长时间,数据库有多少个表 二.技术问题 1 用自己擅长的语言实现非递归单链表反转 现场手写2 Hadoo ...
- C++中的结构体所占内存空间总结
因为结构体有时候需要字节对齐.一般而言,struct 的 sizeof 是所有成员字节对齐后长度相加,而 union 的 sizeof 是取最大的成员长度. 在默认情况下,编译器为每一个变量或数据单元 ...
- Linux 就该这么学 CH03 管道符、重定向和环境变量
0 概述 1 输入输出重定向 重定向技术有5种模式: 标准覆盖输出重定向 标准追加输出重定向 错误覆盖输出重定向 错误追加输出重定向 输入重定向 输入重定向是将文件导入到命令中. 输出重定向是将输入到 ...
- linux 程序失败自动重启
最近写了一个spark streaming 程序,但是程序跑着跑着就报错了,而且不会自动重启,以下脚本实现了程序失败自动重启 基本原理:查看程序日志文件是否有ERROR或Exception字样,有说明 ...
- cube-ui 重构饿了吗Webapp的 scroll-nav域名插槽问题
Vue2.6 将 slot-scope 废弃了. 推荐使用 v-slot: 其使用方法大致如下: 注意多个插槽的情况下,最好都基于 <template> default插槽用法还是一样的, ...
- prototype和__proto__的关系是什么？
我们创建的每个函数都有一个prototype(原型)属性,这个属性是一个对象,它的用途是包含可以由特定类型的所有实例共享的属性和方法.(这个对象下面有个属性,这个属性是另外一个对象的应用 ,这个属性就 ...
- 90% 前端开发者都不知道的 JavaScript 实用小技巧
面试神器之数组去重 const a = [...new Set([1, 2, 3, 3])] >> [1, 2, 3] 操作数组担心 falsy 值? const res = myArra ...
- Linux nfs服务讲解
nfs服务介绍 nfs(Network File System) 网络文件系统,能使用户访问服务器的文件系统,就像访问自己的本机的文件系统一样,并且多个 客户端共享访问该文件系统. 目前nfs服务,较 ...
- Django框架第九篇--Django和Ajax、序列化组件(serializers)、自定义分页器、模型表choice参数
Django和Ajax 一.什么是Ajax AJAX(Asynchronous Javascript And XML)翻译成中文就是“异步的Javascript和XML”.即使用Javascript语 ...