K8s使用说明：

安装方法：将dbackup3-agent-k8s-version.x86_64.tar包传到所有K8S集群节点，
docker:
通过 docker load -i dbackup3-agent-k8s-version.x86_64.tar 载入镜像（每个集群节点均需要载入镜像，载入后可通过 docker image ls 命令查询）

Container:
通过 ctr -n=k8s.io image import dbackup3-agent-k8s-version.x86_64.tar 载入镜像（每个集群节点均需要载入镜像，载入后可通过 ctr -n=k8s.io image ls 命令查询）


上传完成后需要通过以k8s的pod资源方式进行部署和配置，详细部署请参考K8S用户指南。


参考部署配置如下(需根据实际情况进行配置，以下操作均在master节点操作)：

一、 创建命名空间和指定hostid
1）创建dbackup3命名空间
kubectl create namespace dbackup3


2）创建hostid记录该值，在配置dbackup3-k8s-agent.yaml文件的“- name: HOSTID_0”时会使用到

在宿主机执行以下命令：
uuidgen -r | sed "s/-//g"  


3）创建pod完成cluster-admin授权配置

touch cluster.yaml


# cat cluster.yaml 
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: dbackupk8s
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
- kind: ServiceAccount
  name: default
  namespace: dbackup3
  
#执行下面命令，创建资源
kubectl apply -f cluster.yaml  




  
4）sc-config-map.yaml配置文件

touch sc-config-map.yaml

数组当中填写的是Storage Class的名字，根据sc的特性填写到具体位置。

apiVersion: v1
kind: ConfigMap
metadata:
  name: sc-config
  namespace: dbackup3
data:
  # 存储类配置Json文件
  config.json: |
    {
      "csi": [
        "csi-rbd-sc" ,      #注意修改，用kubectl get sc查看，然后根据sc的特性填写
        "csi-sc"        
      ],
      "nfs":[
        "nfs-client"         #注意修改
      ],
      "local": [
        "fast-disks"         #注意修改
      ]
     }

#执行下面命令，创建资源
kubectl apply -f sc-config-map.yaml


5）创建迪备代理pod

touch dbackup3-k8s-agent.yaml

###dbackup3-k8s-agent.yaml 配置参考###

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: dbackup-agent
  namespace: dbackup3
spec:       
  selector:
    matchLabels:
      app: dbackup-agent
  serviceName: dbackup-agent
  replicas: 1
  template:
    metadata:
      labels:
        app: dbackup-agent
    spec:
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            - labelSelector:
                matchLabels:
                  app: dbackup-agent
              topologyKey: "kubernetes.io/hostname" 
      containers:
        - env:
            - name: BACKUPD_HOST
              value: 172.16.30.162                       #迪备，注意修改
            - name: BACKUPD_PORT
              value: "50305"
            - name: HOSTID_0
              value: 6eabbc1a786f49eb8053291e2748ce74
#            - name: HOSTID_1                              #设置多agent时开启
#              value: 8dc565f18be34da3a22eaf17b6b3c399
#            - name: HOSTID_2
#              value: daahwio1309123j21124e129fqwqa11r
            - name: BACKUPD_SSL
              value: "false"
            - name: POD_IMAGE
              value: registry.aliyuncs.com/k8sxio/pause:3.2     #通过docker images | grep pause查看进行更改
            - name: POD_NAME
              value: backup-pod
            - name: DEPLOY_METHOD
              value: statefulset
            - name: HOSTNAME
              valueFrom:
                fieldRef:
                  fieldPath: metadata.name
#            - name: BACKUPD_ACCESS_KEY
#              value: "f1528cdec9afabe99b804e47e0833848"
            - name: BACKUP_NODE            
              valueFrom:
                fieldRef:
                  fieldPath: spec.nodeName
          image: registry.docker.scutech.com/dbackup/stable/focal/dbackup3-agent-k8s:version     #注意修改
          imagePullPolicy: IfNotPresent
          resources:
          name: agent
          securityContext:
            privileged: true
          volumeMounts:
            - mountPath: /var/log/dbackup3
              name: log-volume
            - mountPath: /var/opt/scutech/dbackup3/agent
              name: opt-volume
            - mountPath: /var/lib/kubelet/pods
              mountPropagation: HostToContainer
              name: pods-path
            - mountPath: /dev
              mountPropagation: HostToContainer
              name: dev
            - mountPath: /opt/storage_class
              name: storage-class
            - name: host-time   ###设定本地服务器时区
              mountPath: /etc/localtime
      hostIPC: true
      hostNetwork: true
      hostPID: true
#      nodeName: k8s-master-200          #指定node节点执行
#      nodeSelector:
#          kubernetes.io/hostname: k8s-master-200
      volumes:
        - hostPath:
            path: /var/lib/kubelet/pods
          name: pods-path
        - hostPath:
            path: /dev
          name: dev
        - hostPath:
            path: /opt/data/opt_volume
          name: opt-volume
        - hostPath:
            path: /opt/data/log_volume
          name: log-volume
        - name: host-time
          hostPath: 
            path: /etc/localtime
        - name: storage-class
          configMap:
             name: sc-config
             items: 
             - key: "config.json"
               path: "config.json"
               
               
#执行下面命令，创建资源          
kubectl apply -f dbackup3-k8s-agent.yaml

4）备份代理pod 创建后，可通过 kubectl get pod -n dbackup3 查询。pod 状态为running 即可在web 界面继续进行备份操作。

二、检查kubectl 状态可使用如下命令，备份前可要求管理员进行检查
1）#systemctl status kubelet，查看 kubelet 服务是否为正常 active(running)
2)#systemctl status docker，查看 docker 服务是否为正常 active(running)

3)#kubectl get nodes -owide，查看 kubernetes 集群的部署版本，请确保所有节点的 version 不低于 v1.17.0，否则不支持 CSI 驱动，并查看所有 node 节点是否为 Ready。

4)#kubectl get pod -A，查看 备份资源状态是否为running。