Kubernetes ---- 存储卷(emptyDir、hostPath、NFS)

存储卷

  Pod是有生命周期的,当Pod出现故障时,数据会随着Pod的终结就结束了.
  针对K8s集群,我们应该使用脱离节点的存储设备,共享存储设备.
  如果使用docker持久化数据的那种方法,那么Pod重构后就不能更换节点,否则,挂载的目录位置肯定就访问不到了.

可用存储卷
  1. emptyDir: 临时存储目录,随着Pod删除也会被删除.
  2. hostPath: 节点路径,直接在宿主机上找一个目录,与Pod进行关联.
网络存储卷:
  脱离主机节点本地的存储设备:
  1. SAN: ISCSI,...
  2. NAS: NFS、CIFS
分布式存储:
  1. GlusterFS
  2. cephfs: ceph文件系统存储
  3. rbd: ceph块存储
云端存储:
  1. EBS(AWS): 弹性块存储
  2. Azure Disk(微软)

# 查看k8s支持的存储类型
$ kubectl explain pods.spec.volumes

pods.spec.volumes.emptyDir <Object>
  emptyDir类型的volume在pod分配到node上时被创建,kubernetes会在node上自动分配一个目录,因此无需指定宿主机node上对应的目录文件。这个目录的初始内容为空,
当Pod从node上移除时,emptyDir中的数据会被永久删除.

# 以下为了演示同一Pod的容器可以共享存储空间,虽然挂载目录指定的不一样,但是创建数据最终会在各个定义的目录中出现,实则是一样的.

$ vim pod-vol-demo.yaml
  apiVersion: v1
  kind: Pod
  metadata:
    name: pod-vol-demo
    namespace: default
  spec:
    containers:
    - name: myapp
      image: ikubernetes/myapp:v1
      imagePullPolicy: IfNotPresent
      ports:
      - name: http
        containerPort: 80
      volumeMounts:
      - name: disk1
        mountPath: /usr/share/nginx/html/
    - name: busybox
      image: busybox:latest
      imagePullPolicy: IfNotPresent
      volumeMounts:
      - name: disk1
        mountPath: /data/
      command: ["/bin/sh","-c","while true;do echo $(date) >> /data/index.html;done"]
    volumes:
    - name: disk1
      emptyDir: {}
$ kubectl apply -f pod-vol-demo.yaml
$ kubectl get pods -o wide
NAME     READY STATUS RESTARTS   AGE   IP     NODE   NOMINATED NODE READINESS GATES
pod-vol-demo 2/2 Running   0     43m 10.244.2.124 node2   <none>   <none>

$ curl 10.244.2.124

pods.spec.volumes.hostPath <Object> 宿主机路径
  把Pod所在的宿主机之上的宿主机的文件系统目录与Pod建立关联关系,在Pod被删除时,这个目录是不会被删除的,所以在Pod被删除之后重新被调度到这个节点上的话
对应的数据依然是存在的,如果宿主机上的目录默认不存在的话是否自动创建取决于type的定义,但如果节点宕机了,那么数据还是会丢失的.

path <string> -required-
type <string>

         空字符串(默认)用于向后兼容,这意味着在安装 hostPath 卷之前不会执行任何检查。
DirectoryOrCreate 如果在给定路径上什么都不存在,那么将根据需要创建空目录,权限设置为 0755,具有与 Kubelet 相同的组和所有权。
Directory     在给定路径上必须存在的目录。
FileOrCreate    如果在给定路径上什么都不存在,那么将在那里根据需要创建空文件,权限设置为 0644,具有与 Kubelet 相同的组和所有权。
File        在给定路径上必须存在的文件。
Socket       在给定路径上必须存在的 UNIX 套接字。
CharDevice     在给定路径上必须存在的字符设备。
BlockDevice    在给定路径上必须存在的块设备。

$ vim pod-hostpath.yaml
  apiVersion: v1   kind: Pod   metadata:    name: pod
-hostpath    namespace: default   spec:    containers:    - name: hostpath-container    image: ikubernetes/myapp:v1    imagePullPolicy: IfNotPresent    ports:    - name: http    containerPort: 80    volumeMounts:    - name: disk2    mountPath: /usr/share/nginx/html/    lifecycle:    postStart:    exec:    command: ["/bin/sh","-c","echo $(hostname) > /usr/share/nginx/html/index.html"]    volumes:    - name: disk2    hostPath:    path: /data/pod/volume/pod-hostpath    type: DirectoryOrCreate

$ kubectl apply -f pod-hostpath.yaml
$ kubectl get pods -o wide 
NAME        READY   STATUS RESTARTS AGE     IP    NODE NOMINATED NODE READINESS GATES
pod-hostpath   1/1   Running   0   4m34s 10.244.2.127 node2 <none>   <none>


$ curl 10.244.2.127
pod-hostpath

k8s配置NFS共享存储:

1. 在指定一台服务器安装NFS服务以提供共享存储服务

nfs node:

  # yum -y install nfs-utils
  # mkdir /data/volumes -pv
  # vim /etc/exports
  /data/volumes 192.168.222.0/24(rw,no_root_squash)
  # systemctl start nfs
  # ss -nlt | grep 2049


2. 手动测试k8s的node是否支持挂载NFS文件系统,如果不支持则进行解决.
node2:

  # mount -t nfs 192.168.222.103:/data/volumes/ /mnt

  mount: wrong fs type, bad option, bad superblock on 192.168.222.103:/data/volumes,
  missing codepage or helper program, or other error
  (for several filesystems (e.g. nfs, cifs) you might
  need a /sbin/mount.<type> helper program)

  In some cases useful info is found in syslog - try
  dmesg | tail or so.
  # yum -y install nfs-utils
  # mount -t nfs 192.168.222.103:/data/volumes /mnt
  # mount | grep 192.168.222.103
  # umount /mnt

3. 定义Pod进行测试

$ vim pod-nfs-demo.yaml

  apiVersion: v1   kind: Pod   metadata:    name: nfs
-pod    namespace: default   spec:    containers:    - name: nfs-container    image: ikubernetes/myapp:v1    imagePullPolicy: IfNotPresent    ports:    - name: http    containerPort: 80    volumeMounts:    - name: nfs-disk    mountPath: /usr/share/nginx/html/    lifecycle:    postStart:    exec:    command: ["/bin/sh","-c","echo $(hostname) > /usr/share/nginx/html/index.html"]    volumes:    - name: nfs-disk    nfs:    path: /data/volumes/    server: 192.168.222.103

$ kubectl apply -f pod-nfs-demo.yaml


$ kubectl get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
pod-nfs-vol 1/1 Running 0 4m41s 10.244.2.129 node2 <none> <none>


$ curl 10.244.2.129
pod-nfs-vol

 

3.2 定义Deployment多个Pod同时挂载至NFS
$ vim deploy-nfs-demo.yaml

apiVersion: v1
Kind: Service
metadata:
  name: nfs-svc
  namespace: default
spec:
  selector:
    app: nfs-pod
    release: dev
  type: NodePort
  ports:
  - name: http
    port: 80
    targetPort: 80
    nodePort: 30180
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nfs-deploy
  namespace: default
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nfs-pod
      release: dev
  template:
    meatadata:
      name: nfs-pod
      namespace: defualt
      labels:
        app: nfs-pod
        release: dev
    spec:
      containers:
     - name: nfs-container
       image: ikubernetes/myapp:v2
       imagePullPolicy: IfNotPresent
       ports:
       - name: http
         port: 80
         containerPort: 80
       volumeMounts:
       - name: nfs-disk
         mountPath: /usr/share/nginx/html/
       lifecycle:
         postStart:
           exec:
             command: ["/bin/sh","-c","echo $(hostname) > /usr/share/nginx/html/index.html"]
       livenessProbe:
         exec:
           command: ["/bin/sh","-c","ps aux | grep nginx"]
         initialDelaySeconds: 2
         periodSeconds: 3
       readinessProbe:
         tcpSocket:
         port: 80
       initialDelaySeconds: 2
       periodSeconds: 3
     volumes:
     - name: nfs-disk
       nfs:
         path: /data/volumes/
         server: 192.168.222.103

$ kubectl apply -f deploy-nfs-demo.yaml
$ kubectl get svc -o wide
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE SELECTOR
nfs-svc NodePort 10.101.149.48 <none> 80:30180/TCP 12m app=nfs-pod,release=dev


客户端访问"http://192.168.222.101:30180"

4. 登录至nfs node查看指定目录下是否存在文件

# cat /data/volumes/index.html
pod-nfs-vol

 

posted @ 2020-06-18 09:44  k-free  阅读(1030)  评论(0编辑  收藏  举报