Volume常用的几种卷：
1.emptyDir：本地临时卷：
2.hostPath：本地卷
3.NFS：共享卷
4.configmap: 配置文件
emptyDir: 当 Pod 被分配给节点时，首先创建 emptyDir 卷，并且只要该 Pod 在该节点上运行，该卷就会存在。正如卷的名字所述，它最初是空的。
Pod 中的容器可以读取和写入 emptyDir 卷中的相同文件，尽管该卷可以挂载到每个容器中的相同或不同路径上。当出于任何原因从节点中删除或重构Pod时，emptyDir 中的数据将被永久删除。
[root@localhost7C case3]# cat deploy_empty.yml
#apiVersion: extensions/v1beta1
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
namespace: linux39
spec:
replicas: 1
selector:
matchLabels:
app: ng-deploy-80
template:
metadata:
labels:
app: ng-deploy-80
spec:
containers:
- name: ng-deploy-80
image: nginx
ports:
- containerPort: 80
volumeMounts: #容器卷
- mountPath: /cache #容器路径 ，会自动创建
name: cache-volume #调用宿主机名
volumes: #node节点宿主机卷
- name: cache-volume #node节点宿主机名,也是物理路径名称，会自动创建
emptyDir: {} #启用emptyDir功能
#测试方法 在容器中创建一个文件，在节点上查找(/var/lib/kubelet/pods/) 。
[root@localhost7F k8s]# kubectl apply -f deploy_empty.yml
deployment.apps/nginx-deployment created
[root@localhost7F k8s]# kubectl get pod -A -o wide
linux39 nginx-deployment-7cc86d98d5-csnx8 1/1 Running 0 47s 10.20.2.4 localhost7f.localdomain <none> <none>
[root@localhost7F k8s]#
[root@localhost7F k8s]# kubectl exec -it -n linux39 nginx-deployment-7cc86d98d5-csnx8 bash
root@nginx-deployment-7cc86d98d5-csnx8:/# echo xxxxxx >> /cache/file.txt
root@nginx-deployment-7cc86d98d5-csnx8:/# ls /cache/
file.txt
#宿主机
[root@localhost7J ~]# find /var/lib/kubelet/ -name file.txt
/var/lib/kubelet/pods/ce59b450-18b6-423b-8d1f-03afce8b8c54/volumes/kubernetes.io~empty-dir/cache-volume/file.txt
[root@localhost7J ~]# cat /var/lib/kubelet/pods/ce59b450-18b6-423b-8d1f-03afce8b8c54/volumes/kubernetes.io~empty-dir/cache-volume/file.txt
xxxxxx
[root@localhost7F k8s]# kubectl delete -f deploy_empty.yml
deployment.apps "nginx-deployment" deleted
#文件没有
[root@localhost7J ~]# ll /var/lib/kubelet/pods/
hostPath：hostPath 卷将主机节点的文件系统中的文件或目录挂载到集群中，pod删除的时候，卷不会被删除
#apiVersion: extensions/v1beta1
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
replicas: 1
selector:
matchLabels:
app: ng-deploy-80
template:
metadata:
labels:
app: ng-deploy-80
spec:
containers:
- name: ng-deploy-80
image: nginx
ports:
- containerPort: 80
volumeMounts: #容器卷
- mountPath: /data/mysql #容器路径 ，会自动创建
name: data-volume #调用宿主机名
volumes: #node节点宿主机卷
- name: data-volume #node节点宿主机名
hostPath: #启用hostpath功能
path: /data/mysql #node节点物理路径名称，会自动创建
#测试方法 在容器中创建一个文件，在节点上查找（/var/lib/kubelet/pods/） 。
#思考：如果pod重构后在另一个节点，数据文件会复制过去吗？答案是不能。
#思考：根据第一个问题，文件是不复制过去， 可以使用nodeSelector节点筛选器，使pod只运行某一节点。还有一方案是使用NFS共享卷。
[root@localhost7C k8s]# kubectl apply -f deploy_hostPath.yml
deployment.apps/nginx-deployment created
[root@localhost7C k8s]# kubectl get pod -A -o wide
NAMESPACE NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
default nginx-deployment-7944748bc4-f9vvl 1/1 Running 0 34s 10.20.2.5 localhost7f.localdomain <none> <none>
[root@localhost7C k8s]C kubectl exec -it nginx-deployment-7944748bc4-f9vvl bash
root@nginx-deployment-7944748bc4-f9vvl:/# echo aaaa >> /data/mysql/file.txt
#文件在
[root@localhost7F~]# cat /data/mysql/file.txt
aaaa
#drain命令：用于在维护期间排除节点,驱逐node上的业务pod，用于node节点下线。
[root@localhost7C k8s]# kubectl drain localhost7f.localdomain
[root@localhost7C case3]# kubectl get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-deployment-7944748bc4-dn7ss 1/1 Running 0 96s 10.10.4.28 localhost7g.localdomain <none> <none>
#没有文件
[root@localhost7G ~]# ll /data/mysql/
[root@localhost7C k8s]C kubectl exec -it nginx-deployment-7944748bc4-dn7ss bash
root@nginx-deployment-7944748bc4-f9vvl:/# echo aaaa >> /data/mysql/file.txt
[root@localhost7C k8s]# kubectl delete -f deploy_hostPath.yml
deployment.apps "nginx-deployment" deleted
#pod删除文件还在
[root@localhost7J ~]# cat /data/mysql/file.txt
aaaa
NFS: nfs卷允许将现有的 NFS（网络文件系统）共享挂载到您的容器中。不像 emptyDir，当删除 Pod 时，nfs 卷的内容被保留，卷仅仅是
被卸载。这意味着 NFS 卷可以预填充数据，并且可以在 pod 之间“切换”数据。 NFS 可以被多个写入者同时挂载。
#apiVersion: extensions/v1beta1
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
replicas: 1
selector:
matchLabels:
app: ng-deploy-80
template:
metadata:
labels:
app: ng-deploy-80
spec:
containers:
- name: ng-deploy-80
image: nginx
ports:
- containerPort: 80
volumeMounts: #启用卷挂载
- mountPath: /usr/share/nginx/html/mysite #容器挂载路径,mysite文件夹会自动创建。
name: my-nfs-volume #卷名，调用NFS共享名
- mountPath: /data/nginx/html
name: linux39-nfs-volume
volumes:
- name: my-nfs-volume #随机的NFS共享名
nfs:
server: 192.168.80.110 #NFS地址
path: /nfs/datanfs #NFS路径名称,先创建。
- name: linux39-nfs-volume
nfs:
server: 192.168.80.110
path: /data/linux39
---
apiVersion: v1
kind: Service
metadata:
name: ng-deploy-80
spec:
ports:
- name: http
port: 81
targetPort: 80
nodePort: 30016
protocol: TCP
type: NodePort
selector:
app: ng-deploy-80
#查看下node宿主机的挂载情况。
#测试中发现，多个pod可以同时挂载同一个NFS
#测试中发现，NFS的地址范围除了pod地址，也要包括node节点的地址，因为NFS挂载是先挂载node宿主机，再node宿主机挂载到pod中的。
#安装NFS
[root@localhost7B ~]#yum install nfs-utils.x86_64 -y
[root@localhost7B ~]#systemctl start nfs.service
[root@localhost7B ~]# mkdir /nfs/datanfs -p
[root@localhost7B ~]# vim /etc/exports
/nfs/datanfs *(rw)
[root@localhost7B ~]# systemctl restart nfs.service
[root@localhost7B ~]# echo nginx config file > /nfs/datanfs/nginx.conf
#另一机子测试
showmount -e 192.168.80.110
mount 192.168.80.100:/nfs/datanfs /mnt/cdrom/
umount /mnt/cdrom/
[root@localhost7F k8s]# kubectl apply -f deploy_nfs.yml
deployment.apps/nginx-deployment created
service/ng-deploy-80 created
[root@localhost7F k8s]# kubectl get pod -A
NAMESPACE NAME READY STATUS RESTARTS AGE
default nginx-deployment-76567bccb4-zg765 1/1 Running 0 64s
[root@localhost7F k8s]# kubectl exec -it nginx-deployment-76567bccb4-zg765 bash
root@nginx-deployment-76567bccb4-zg765:/# ll /usr/share/nginx/html/
50x.html index.html mysite/
root@nginx-deployment-76567bccb4-zg765:/# cat /usr/share/nginx/html/mysite/nginx.conf
nginx config file
#容器的挂载信息
root@nginx-deployment-76567bccb4-zg765:/# mount | grep datanfs
192.168.80.110:/nfs/datanfs on /usr/share/nginx/html/mysite type nfs4 (rw,relatime,vers=4.1,rsize=262144,wsize=262144,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=192.168.80.150,local_lock=none,addr=192.168.80.110)
#宿主机的挂载信息
[root@localhost7F ~]# mount | grep datanfs
192.168.80.110:/nfs/datanfs on /var/lib/kubelet/pods/bdae2519-de34-4219-8314-38caad08673a/volumes/kubernetes.io~nfs/my-nfs-volume type nfs4 (rw,relatime,vers=4.1,rsize=262144,wsize=262144,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=192.168.80.150,local_lock=none,addr=192.168.80.110)
configmap:将配置信息放到configmap对象中，然后在pod的对象中导入configmap对象，实现导入配置的操作
声明一个ConfigMap的对象，作为Volume挂载到pod中
#apiVersion: extensions/v1beta1
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
replicas: 1
selector:
matchLabels:
app: ng-deploy-80
template:
metadata:
labels:
app: ng-deploy-80
spec:
containers:
- name: ng-deploy-80
image: nginx
ports:
- containerPort: 80
volumeMounts: #启用卷挂载
- mountPath: /usr/share/nginx/html/mysite #容器挂载路径,mysite文件夹会自动创建。
name: my-nfs-volume #卷名，调用NFS共享名
- mountPath: /data/nginx/html
name: linux39-nfs-volume
volumes:
- name: my-nfs-volume #随机的NFS共享名
nfs:
server: 192.168.80.110 #NFS地址
path: /nfs/datanfs #NFS路径名称,先创建。
- name: linux39-nfs-volume
nfs:
server: 192.168.80.110
path: /data/linux39
---
apiVersion: v1
kind: Service
metadata:
name: ng-deploy-80
spec:
ports:
- name: http
port: 81
targetPort: 80
nodePort: 30016
protocol: TCP
type: NodePort
selector:
app: ng-deploy-80
#查看下node宿主机的挂载情况。
#测试中发现，多个pod可以同时挂载同一个NFS
#测试中发现，NFS的地址范围除了pod地址，也要包括node节点的地址，因为NFS挂载是先挂载node宿主机，再node宿主机挂载到pod中的。
#安装NFS
[root@localhost7B ~]#yum install nfs-utils.x86_64 -y
[root@localhost7B ~]#systemctl start nfs.service
[root@localhost7B ~]# mkdir /nfs/datanfs -p
[root@localhost7B ~]# vim /etc/exports
/nfs/datanfs *(rw)
[root@localhost7B ~]# systemctl restart nfs.service
[root@localhost7B ~]# echo nginx config file > /nfs/datanfs/nginx.conf
#另一机子测试
showmount -e 192.168.80.110
mount 192.168.80.100:/nfs/datanfs /mnt/cdrom/
umount /mnt/cdrom/
[root@localhost7F k8s]# kubectl apply -f deploy_nfs.yml
deployment.apps/nginx-deployment created
service/ng-deploy-80 created
[root@localhost7F k8s]# kubectl get pod -A
NAMESPACE NAME READY STATUS RESTARTS AGE
default nginx-deployment-76567bccb4-zg765 1/1 Running 0 64s
[root@localhost7F k8s]# kubectl exec -it nginx-deployment-76567bccb4-zg765 bash
root@nginx-deployment-76567bccb4-zg765:/# ll /usr/share/nginx/html/
50x.html index.html mysite/
root@nginx-deployment-76567bccb4-zg765:/# cat /usr/share/nginx/html/mysite/nginx.conf
nginx config file
#容器的挂载信息
root@nginx-deployment-76567bccb4-zg765:/# mount | grep datanfs
192.168.80.110:/nfs/datanfs on /usr/share/nginx/html/mysite type nfs4 (rw,relatime,vers=4.1,rsize=262144,wsize=262144,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=192.168.80.150,local_lock=none,addr=192.168.80.110)
#宿主机的挂载信息
[root@localhost7F ~]# mount | grep datanfs
192.168.80.110:/nfs/datanfs on /var/lib/kubelet/pods/bdae2519-de34-4219-8314-38caad08673a/volumes/kubernetes.io~nfs/my-nfs-volume type nfs4 (rw,relatime,vers=4.1,rsize=262144,wsize=262144,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=192.168.80.150,local_lock=none,addr=192.168.80.110)
实例2：将ConfigMap中的数据设置为容器的环境变量
apiVersion: v1
kind: ConfigMap
#定义configmap名称，
metadata:
name: nginx-config
data:
username: user1 #pod中变量的值,可以定义多个。
---
#apiVersion: extensions/v1beta1
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
replicas: 1
selector:
matchLabels:
app: ng-deploy-80
template:
metadata:
labels:
app: ng-deploy-80
spec:
containers:
- name: ng-deploy-80
image: nginx
env:
- name: MY_USERNAME #pod中定义的变量名
valueFrom:
configMapKeyRef: #
name: nginx-config #调用configmap名称
key: username
ports:
- containerPort: 80
[root@localhost7F k8s]# kubectl apply -f deploy_configmap.yml
root@nginx-deployment-669d88577-vf4nh:/# kubectl exec -it nginx-deployment-669d88577-vf4nh bash
root@nginx-deployment-669d88577-vf4nh:/# echo $MY_USERNAME
user1
标签：说明,name,deploy,Volume,nginx,nfs,deployment,k8s,root From： https://www.cnblogs.com/Yuanbangchen/p/17188564.html