Kubernets使用glusterfs存储
参考：
https://www.cnblogs.com/breezey/p/8849466.html
https://www.cnblogs.com/zhaojiedi1992/p/zhaojiedi_liunx_54_kubernates_glusterfs-heketi.html
https://www.cnblogs.com/panwenbin-logs/p/10231859.html
https://www.cnblogs.com/huangyanqi/p/8406534.html
k8s中部署Heketi和Gluster 一：https://blog.51cto.com/newfly/2134514
k8s中部署Heketi和Gluster 二：https://blog.51cto.com/newfly/2139393
环境
| IP地址 | 部署程序 |
|---|---|
| 192.168.10.11 | GlusterFS,Heketi |
| 192.168.10.12 | GlusterFS |
| 192.168.10.13 | GlusterFS |
Gluster安装
# yum 安装Gluster
yum install -y centos-release-gluster
yum install -y glusterfs glusterfs-server glusterfs-fuse glusterfs-rdma
#启动Gluster
systemctl start glusterd && systemctl enable glusterd && systemctl status glusterd
在任意节点发现其他节点
gluster peer probe 192.168.10.12
gluster peer probe 192.168.10.13
gluster peer status #查看集群状态
Number of Peers: 2
Hostname: 192.168.10.12
Uuid: 80d5879f-ba7a-4a9b-817c-9ebb05d7086e
State: Peer in Cluster (Connected)
Hostname: 192.168.10.13
Uuid: 7b0f53c9-296b-4924-b347-3dabfc224118
State: Peer in Cluster (Connected)
安装Heketi
yum install heketi heketi-client -y
# 配置heketi用户能够基于SSH秘钥认证的方式连接至GlusterFS集群各节点
ssh-keygen -f /etc/heketi/heketi_key -t rsa -N ''
chown heketi:heketi /etc/heketi/heketi*
ssh-copy-id -i /etc/heketi/heketi_key.pub root@192.168.10.11
ssh-copy-id -i /etc/heketi/heketi_key.pub root@192.168.10.12
ssh-copy-id -i /etc/heketi/heketi_key.pub root@192.168.10.13
#验证：ssh -i /etc/heketi/heketi_key root@192.168.10.13
# heketi的主配置文件/etc/heketi/heketi.json，定义服务监听的端口、认证及连接Gluster存储集群的方式
cat <<EOF> /etc/heketi/heketi.json
{
"port": "18080",
"use_auth": false,
"jwt": {
"admin": {
"key": "adminSecret"
},
"user": {
"key": "userSecret"
}
},
"glusterfs": {
"executor": "ssh",
"sshexec": {
"keyfile": "/etc/heketi/heketi_key",
"user": "root",
"port": "22",
"fstab": "/etc/fstab"
},
"db": "/var/lib/heketi/heketi.db",
"loglevel" : "debug"
}
}
EOF
#启动heketi服务
systemctl start heketi && systemctl enable heketi && systemctl status heketi
# 测试
curl http://192.168.10.11:18080/hello
# 返回：Hello from Heketi
设置Heketi拓扑
指定具体的节点定义在同一个集群中，并指明各节点用于提供存储的设备
cat <<EOF> /etc/heketi/topolgy_demo.json
{
"clusters": [
{
"nodes": [
{
"node": {
"hostnames": {
"manage": [
"192.168.10.11"
],
"storage": [
"192.168.10.11"
]
},
"zone": 1
},
"devices": [
"/dev/sdb"
]
},
{ "node": {
"hostnames": {
"manage": [
"192.168.10.12"
],
"storage": [
"192.168.10.12"
]
},
"zone": 1
},
"devices": [
"/dev/sdb"
]
},
{ "node": {
"hostnames": {
"manage": [
"192.168.10.13"
],
"storage": [
"192.168.10.13"
]
},
"zone": 1
},
"devices": [
"/dev/sdb"
]
}
]
}
]
}
EOF
#加载拓扑信息
export HEKETI_CLI_SERVER=http://192.168.10.11:18080
echo $HEKETI_CLI_SERVER
heketi-cli topology load --json=topolgy_demo.json
#显示如下
Creating cluster ... ID: d1475a413956ce3187ad83299ec39ef7
Allowing file volumes on cluster.
Allowing block volumes on cluster.
Creating node 192.168.10.11 ... ID: 8ab4d5f236904f353c8acd4a7a721a76
Adding device /dev/sdb ... OK
Creating node 192.168.10.12 ... ID: 82377f431888e99ac2a2460257ee97a6
Adding device /dev/sdb ... OK
Creating node 192.168.10.13 ... ID: 8dc8e9eb476f23c49d7f22b78a660cc6
Adding device /dev/sdb ... OK
# 根据生产的随机Cluster ID 查看集群状态
heketi-cli cluster info d1475a413956ce3187ad83299ec39ef7
#显示如下
Cluster id: d1475a413956ce3187ad83299ec39ef7
Nodes:
82377f431888e99ac2a2460257ee97a6
8ab4d5f236904f353c8acd4a7a721a76
8dc8e9eb476f23c49d7f22b78a660cc6
Volumes:
Block: true
File: true
测试
# 创建测试使用的存储卷
heketi-cli volume create --size=5
##显示如下
Cluster id: d1475a413956ce3187ad83299ec39ef7
Nodes:
82377f431888e99ac2a2460257ee97a6
8ab4d5f236904f353c8acd4a7a721a76
8dc8e9eb476f23c49d7f22b78a660cc6
Volumes:
Block: true
File: true
[root@kube11 glusterd]# heketi-cli volume create --size=5
Name: vol_ada73b178c205db2bea66bca91d578c3
Size: 5
Volume Id: ada73b178c205db2bea66bca91d578c3
Cluster Id: d1475a413956ce3187ad83299ec39ef7
Mount: 192.168.10.12:vol_ada73b178c205db2bea66bca91d578c3 #客户端挂载的地址
Mount Options: backup-volfile-servers=192.168.10.11,192.168.10.13
Block: false
Free Size: 0
Reserved Size: 0
Block Hosting Restriction: (none)
Block Volumes: []
Durability Type: replicate
Distribute Count: 1
Replica Count: 3
# 客户端挂载存储卷
yum install centos-release-gluster -y
yum install glusterfs glusterfs-fuse -y
mount -t glusterfs 192.168.10.12:vol_ada73b178c205db2bea66bca91d578c3 /mnt ##创建volume信息中的Mount信息
#也可以写到 /etc/fstab 文件里
df -h #df -h 就可以查到挂载的文件
文件系统 容量 已用 可用 已用% 挂载点
192.168.10.12:vol_ada73b178c205db2bea66bca91d578c3 5.0G 84M 5.0G 2% /mnt
heketi-cli volume delete ada73b178c205db2bea66bca91d578c3 #测试完成后删除存储卷
k8s中使用GlusterFS
k8s中使用glusterfs的时候， 会根据pvc的申请自动创建对应的pv， 然后绑定。 这样我们在pod里面就可以指定具体的pvc了。
#创建一个StorageClass
cat <<EOF> gluster-storage-class.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: glusterfs #-------------存储类的名字
provisioner: kubernetes.io/glusterfs
parameters:
resturl: "http://192.168.10.11:18080" #-------------heketi 的ip 和端口
restuser: "admin" #-------------heketi的认证用户，这里随便填，因为没有启用鉴权模式
gidMin: "40000"
gidMax: "50000"
volumetype: "replicate:3" #-------------申请的默认为3副本模式，因为目前有三个gluster节点。
EOF
kubectl apply -f gluster-storage-class.yaml
# 创建PVC
cat <<EOF> gluster-pvc.yaml
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: gluster1
namespace: myapp
annotations:
volume.beta.kubernetes.io/storage-class: glusterfs #----------上面创建的存储类的名称
spec:
accessModes:
- ReadWriteMany
resources:
requests:
storage: 5Gi
EOF
kubectl apply -f gluster-pvc.yaml
#个pvc对应一个brick，一个brick对应一个LV
kubectl get pvc -n myapp
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
gluster1 Bound pvc-f07c6e74-2790-11ea-bf97-000c29d937a0 5Gi RWX glusterfs 15s
vgs
VG #PV #LV #SN Attr VSize VFree
centos 1 2 0 wz--n- <49.00g 0
vg_800718e6161e0bfbcf2e0105efcdb664 1 2 0 wz--n- 19.87g <14.82g
lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
root centos -wi-ao---- <47.00g
swap centos -wi-ao---- 2.00g
brick_1cb2402e5d74c004cffd40f57f5cb2c7 vg_800718e6161e0bfbcf2e0105efcdb664 Vwi-aotz-- 5.00g tp_dca1f5af68eeb304921c5bcf8f382308 0.24
tp_dca1f5af68eeb304921c5bcf8f382308 vg_800718e6161e0bfbcf2e0105efcdb664 twi-aotz-- 5.00g 0.24 10.14
创建pod验证
# 创建pod验证
cat <<EOF> heketi-nginx.yaml
apiVersion: v1
kind: Pod
metadata:
name: nginx-pod1
namespace: myapp
labels:
name: nginx-pod1
spec:
containers:
- name: nginx-pod1
image: nginx:latest
ports:
- name: web
containerPort: 80
volumeMounts:
- name: gluster-vol1
mountPath: /usr/share/nginx/html
volumes:
- name: gluster-vol1
persistentVolumeClaim:
claimName: gluster1 #上面创建的pvc
EOF
kubectl apply -f heketi-nginx.yaml
kubectl get pod -n myapp -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-pod1 1/1 Running 0 18s 172.30.98.14 192.168.10.11 <none> <none>
kubectl exec -it nginx-pod1 /bin/sh
cd /usr/share/html/
echo '192.168.10.11' >> index.html
touch fana.txt #写个文件，过会，就会在所有glusterfs节点 查看到
# 在其他机器访问
curl http://172.30.98.14
#显示：192.168.10.11
#登录其他glusterfs机器
df -h #会出现/var/lib/heketi/mounts/里挂载的目录
/dev/mapper/vg_6e72843352751e3822bb13b55464f7cc-brick_2434b52740c2a485d057f5f39ca54628 5.0G 33M 5.0G 1% /var/lib/heketi/mounts/vg_6e72843352751e3822bb13b55464f7cc/brick_2434b52740c2a485d057f5f39ca54628
#cd 进去会看到 index.html fana.txt 文件
