k8s PodPreset

1. 在master中加入配置

如果有多个master, 每个master都需要加入

/etc/kubernetes/manifests/kube-apiserver.yaml

- --enable-admission-plugins=NodeRestriction,PodPreset
- --service-node-port-range=1-65535
- --runtime-config=settings.k8s.io/v1alpha1=true

完整的配置文件

https://gitee.com/chen1219_1/k8s-install/raw/master/PodPreset/kube-apiserver.yaml

重启kubelet

systemctl restaet kubelet

2. 测试一个简单的podpreset

为default 名称空间内所有的pod 设置环境变量TZ

mkdir podpreset
cd podpreset
vim test.yaml

apiVersion: settings.k8s.io/v1alpha1
kind: PodPreset
metadata:
  name: setting-timezone
spec:
  selector:
    matchLabels:
  env:
    - name: TZ
      value: Asia/Shanghai

kubectl apply -f test.yaml

kubectl get PodPreset

查看default名称空间内的pod是否有环境变量 TZ

3. 指定pod挂载一个pvc

3.1 搭建NFS服务

192.168.31.200

3.1.1 安装 nfs-utils

yum install nfs-utils -y
systemctl start rpcbind  &&systemctl enable rpcbind
systemctl start nfs && systemctl enable nfs

3.1.2 建立nfs存储目录

echo "/data/kubernetes/ *(rw,no_root_squash,no_all_squash,sync)" >/etc/exports
exportfs -r

Copyecho "/data/kubernetes/ *(rw,no_root_squash,no_all_squash,sync)" >/etc/exports
exportfs -r

3.1.3 查看nfs挂载

exportfs -v

/data/kubernetes
		<world>(sync,wdelay,hide,no_subtree_check,sec=sys,rw,secure,no_root_squash,no_all_squash)

3.1.4 使得配置生效

showmount -e

Export list for hdss200.host.com:
/data/kubernetes *

如果出现报错 clnt_create: RPC: Program not registered

systemctl stop rpcbind
systemctl stop nfs
systemctl start rpcbind
systemctl start nfs

3.2. 在master上创建PV

192.168.31.37

3.2.1 创建rbac

rbac.yaml

https://gitee.com/chen1219_1/k8s-install/raw/master/nfs-client-provisioner/rbac.yaml

kubectl apply -f rbac.yaml

3.2.2 创建sc

storageclass.yaml

CopyapiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: 200-nfs-storage
provisioner: 200-nfs-provisioner     #这里要和第三个nfs-client-provisioner的env环境变量中的value值对应。
reclaimPolicy: Retain

3.3 创建PVC，绑定PV

3.3.1 创建nfs-client-provisioner容器

deployment.yaml

CopyapiVersion: apps/v1
kind: Deployment
metadata:
  name: nfs-client-provisioner
  namespace: kube-system
spec:
  replicas: 1               #副本数量为1
  strategy:
    type: Recreate
  selector:
    matchLabels:
      app: nfs-client-provisioner
  template:
    metadata:
      labels:
        app: nfs-client-provisioner
    spec:
      serviceAccount: nfs-client-provisioner       #指定账户
      containers:
        - name: nfs-client-provisioner
          image: harbor.od.com/public/nfs-client-provisioner:latest   #使用的是这个镜像
          volumeMounts:
            - name: nfs-client-root
              mountPath:  /persistentvolumes      #指定容器内的挂载目录
          env:
            - name: PROVISIONER_NAME        #这是这个容器内置的变量
              value: 200-nfs-provisioner         #这是上面变量的值（名字）
            - name: NFS_SERVER       #内置变量，用于指定nfs服务的IP
              value: 192.168.31.200            
            - name: NFS_PATH              #内置变量，指定的是nfs共享的目录
              value: /data/kubernetes
      volumes:              #这下面是指定上面挂载到容器内的nfs的路径及IP
        - name: nfs-client-root
          nfs:
            server: 192.168.31.200
            path: /data/kubernetes

kubectl apply -f deployment.yaml

3.3.2 创建pvc

test2-pvc.yaml

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: test2-nginx-pvc
  namespace: ingress-nginx        
spec:
  storageClassName: 200-nfs-storage                 #定义存储类的名字，要和SC的名字对应
  accessModes:
    - ReadWriteMany          #访问模式为RWM
  resources:
    requests:
      storage: 500Mi

3.4 创建podPreset

在 ingress-nginx 名称空间中的pod带 role: java-inject 标签的挂载/mymountpath目录， /mymountpath 挂载pvc test2-nginx-pvc

3.4.1 创建PodPreset 资源

addmount.yaml

apiVersion: settings.k8s.io/v1alpha1
kind: PodPreset
metadata:
  name: mountpath
  namespace: ingress-nginx
spec:
  selector:
    matchLabels:
      role: java-inject
  volumeMounts:
  - mountPath: /mymountpath
    name: mymount
  volumes:
  - name: mymount
    persistentVolumeClaim:
      claimName: test2-nginx-pvc

3.4.2 部署PodPreset资源

kubectl apply -f addmount.yaml

3.4.3 部署PodPreset资源

kubectl get PodPreset -n ingress-nginx

3.4.4 在pvc中创建文件