网络存储卷之nfs

搭建 nfs 服务器

搭建 nfs 服务器

1. 安装配置 nfs

[root@k8smaster001 nfs-pvc]# yum -y install nfs-utils rpcbind -y

2. 创建 /nfs-data/nfs 为 nfs 数据目录

[root@k8smaster001 nfs-pvc]# mkdir /nfs-data/nfs -p

3. 默认配置文件 /etc/exports 下，在该文件中添加下面的配置信息

vim /etc/exports
/nfs-data/nfs       *(rw,no_root_squash,sync)

4. 启动 nfs 服务

systemctl restart nfs-server && systemctl enable nfs-server

5. 所有node节点安装nfs客户端相关软件包

yum install nfs-utils -y

6. 验证 fs 可用性

[root@k8sworker001 ~]# showmount -e 192.168.3.201
Export list for 192.168.3.201:
/nfs-data/nfs *

创建应用清单使用 nfs

1. master 节点上创建 volume-nfs.yml 文件

apiVersion: apps/v1
kind: Deployment
metadata:
  name: volume-nfs
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.15-alpine
        imagePullPolicy: IfNotPresent
        volumeMounts:
        - name: documentroot
          mountPath: /usr/share/nginx/html
        ports:
        - containerPort: 80
      volumes:
      - name: documentroot
        nfs:
          server: 192.168.3.201
          path: /nfs-data/nfs/

2. 应用 yaml 创建

[root@k8smaster001 nfs]# kubectl apply -f volume-nfs.yml 
deployment.apps/volume-nfs created

3. 在 nfs 服务器共享目录中创建验证文件

[root@k8smaster001 nfs]# echo "volume-nfs" > /nfs-data/nfs/index.html

4. 在 pod 中验证, 发现可以直接访问到，查看容器内文件也存在

[root@k8smaster002 ~]# kubectl get pod -owide|grep volume-nfs
volume-nfs-74cbcccff6-5b97p   1/1     Running   0          3m13s   10.244.35.43    k8sworker002   <none>           <none>
volume-nfs-74cbcccff6-szwk9   1/1     Running   0          3m13s   10.244.87.206   k8sworker001   <none>           <none>
[root@k8smaster002 ~]# curl 10.244.35.43
volume-nfs
[root@k8smaster002 ~]# curl 10.244.87.206
volume-nfs
[root@k8smaster002 ~]# kubectl exec -it volume-nfs-74cbcccff6-5b97p -- cat /usr/share/nginx/html/index.html
volume-nfs

5. 删除 pod 再次启动，验证文件是否存在

[root@k8smaster001 nfs]# kubectl delete -f volume-nfs.yml 
deployment.apps "volume-nfs" deleted
[root@k8smaster001 nfs]# kubectl apply -f volume-nfs.yml 
deployment.apps/volume-nfs created

6. 发现 容器已经改变、查看文件还存在，证明了 pod 摧毁 文件存在本地

[root@k8smaster002 ~]# kubectl get pod -owide|grep volume-nfs
volume-nfs-74cbcccff6-dctr7   1/1     Running   0          61s   10.244.35.44    k8sworker002   <none>           <none>
volume-nfs-74cbcccff6-tftmh   1/1     Running   0          61s   10.244.87.193   k8sworker001   <none>           <none>
[root@k8smaster002 ~]# curl 10.244.35.44
volume-nfs
[root@k8smaster002 ~]# curl 10.244.87.193
volume-nfs
[root@k8smaster002 ~]# kubectl exec -it volume-nfs-74cbcccff6-dctr7 -- cat /usr/share/nginx/html/index.html
volume-nfs