安装k8s可视化界面（图形化界面）

一，首先安装dasboard

将kubernetes-dashboard需要的镜像上传到工作节点（hd2.com)和hd3.com

将其上传为镜像：docker load -i dashboard_2_0_0.tar.gz(hd2.com,hd3.com)

vim kubernetes-dashboard.yaml

apiVersion: v1
kind: Namespace
metadata:
  name: kubernetes-dashboard

---

apiVersion: v1
kind: ServiceAccount
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kubernetes-dashboard

---

kind: Service
apiVersion: v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kubernetes-dashboard
spec:
  ports:
    - port: 443
      targetPort: 8443
  selector:
    k8s-app: kubernetes-dashboard

---

apiVersion: v1
kind: Secret
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard-certs
  namespace: kubernetes-dashboard
type: Opaque

---

apiVersion: v1
kind: Secret
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard-csrf
  namespace: kubernetes-dashboard
type: Opaque
data:
  csrf: ""

---

apiVersion: v1
kind: Secret
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard-key-holder
  namespace: kubernetes-dashboard
type: Opaque

---

kind: ConfigMap
apiVersion: v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard-settings
  namespace: kubernetes-dashboard

---

kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kubernetes-dashboard
rules:
  # Allow Dashboard to get, update and delete Dashboard exclusive secrets.
  - apiGroups: [""]
    resources: ["secrets"]
    resourceNames: ["kubernetes-dashboard-key-holder", "kubernetes-dashboard-certs", "kubernetes-dashboard-csrf"]
    verbs: ["get", "update", "delete"]
    # Allow Dashboard to get and update 'kubernetes-dashboard-settings' config map.
  - apiGroups: [""]
    resources: ["configmaps"]
    resourceNames: ["kubernetes-dashboard-settings"]
    verbs: ["get", "update"]
    # Allow Dashboard to get metrics.
  - apiGroups: [""]
    resources: ["services"]
    resourceNames: ["heapster", "dashboard-metrics-scraper"]
    verbs: ["proxy"]
  - apiGroups: [""]
    resources: ["services/proxy"]
    resourceNames: ["heapster", "http:heapster:", "https:heapster:", "dashboard-metrics-scraper", "http:dashboard-metrics-scraper"]
    verbs: ["get"]

---

kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
rules:
  # Allow Metrics Scraper to get metrics from the Metrics server
  - apiGroups: ["metrics.k8s.io"]
    resources: ["pods", "nodes"]
    verbs: ["get", "list", "watch"]

---

apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kubernetes-dashboard
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: Role
  name: kubernetes-dashboard
subjects:
  - kind: ServiceAccount
    name: kubernetes-dashboard
    namespace: kubernetes-dashboard

---

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: kubernetes-dashboard
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: kubernetes-dashboard
subjects:
  - kind: ServiceAccount
    name: kubernetes-dashboard
    namespace: kubernetes-dashboard

---

kind: Deployment
apiVersion: apps/v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kubernetes-dashboard
spec:
  replicas: 1
  revisionHistoryLimit: 10
  selector:
    matchLabels:
      k8s-app: kubernetes-dashboard
  template:
    metadata:
      labels:
        k8s-app: kubernetes-dashboard
    spec:
      containers:
        - name: kubernetes-dashboard
          image: kubernetesui/dashboard:v2.0.0-beta8
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 8443
              protocol: TCP
          args:
            - --auto-generate-certificates
            - --namespace=kubernetes-dashboard
            # Uncomment the following line to manually specify Kubernetes API server Host
            # If not specified, Dashboard will attempt to auto discover the API server and connect
            # to it. Uncomment only if the default does not work.
            # - --apiserver-host=http://my-address:port
          volumeMounts:
            - name: kubernetes-dashboard-certs
              mountPath: /certs
              # Create on-disk volume to store exec logs
            - mountPath: /tmp
              name: tmp-volume
          livenessProbe:
            httpGet:
              scheme: HTTPS
              path: /
              port: 8443
            initialDelaySeconds: 30
            timeoutSeconds: 30
          securityContext:
            allowPrivilegeEscalation: false
            readOnlyRootFilesystem: true
            runAsUser: 1001
            runAsGroup: 2001
      volumes:
        - name: kubernetes-dashboard-certs
          secret:
            secretName: kubernetes-dashboard-certs
        - name: tmp-volume
          emptyDir: {}
      serviceAccountName: kubernetes-dashboard
      nodeSelector:
        "beta.kubernetes.io/os": linux
      # Comment the following tolerations if Dashboard must not be deployed on master
      tolerations:
        - key: node-role.kubernetes.io/master
          effect: NoSchedule

---

kind: Service
apiVersion: v1
metadata:
  labels:
    k8s-app: dashboard-metrics-scraper
  name: dashboard-metrics-scraper
  namespace: kubernetes-dashboard
spec:
  ports:
    - port: 8000
      targetPort: 8000
  selector:
    k8s-app: dashboard-metrics-scraper

---

kind: Deployment
apiVersion: apps/v1
metadata:
  labels:
    k8s-app: dashboard-metrics-scraper
  name: dashboard-metrics-scraper
  namespace: kubernetes-dashboard
spec:
  replicas: 1
  revisionHistoryLimit: 10
  selector:
    matchLabels:
      k8s-app: dashboard-metrics-scraper
  template:
    metadata:
      labels:
        k8s-app: dashboard-metrics-scraper
      annotations:
        seccomp.security.alpha.kubernetes.io/pod: 'runtime/default'
    spec:
      containers:
        - name: dashboard-metrics-scraper
          image: kubernetesui/metrics-scraper:v1.0.1
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 8000
              protocol: TCP
          livenessProbe:
            httpGet:
              scheme: HTTP
              path: /
              port: 8000
            initialDelaySeconds: 30
            timeoutSeconds: 30
          volumeMounts:
          - mountPath: /tmp
            name: tmp-volume
          securityContext:
            allowPrivilegeEscalation: false
            readOnlyRootFilesystem: true
            runAsUser: 1001
            runAsGroup: 2001
      serviceAccountName: kubernetes-dashboard
      nodeSelector:
        "beta.kubernetes.io/os": linux
      # Comment the following tolerations if Dashboard must not be deployed on master
      tolerations:
        - key: node-role.kubernetes.io/master
          effect: NoSchedule
      volumes:
        - name: tmp-volume
          emptyDir: {}

刷新资源清单：kubectl appy -f kubernetes-dashboard.yaml

查看状态：kubectl get pods -n kubernetes-dashboard

 

 修改service type类型变成NodePort：

kubectl edit svc kubernetes-dashboard-n kubernetes-dashboard

将type：CulsterIP改为NodePort保存退出，再次进行查看，则可以看到：

原来的cluster变成了NodePort如上图所示。

使用主机IP和上图暴露出的端口即可访问：图形化界面

通过token 令牌访问dashboard：kubectl create clusterrolebinding dashboard-cluster-admin--clusterrole=cluster-admin--serviceaccount=kubernetes-dashboard:kubernetes-dashboard

查看kubectl-dashboard名称空间下的secret：kubectl get secret -n kubernetes-dashboard

找到对应的带有token值：kubectl describe secret 和带有token的kubernetes-dashboard名。会产生一个令牌。

将此令牌复制到访问界面的token下，点击login即可登录

通过kubernetes-dashboard创建容器。

首先将nginx.tar.gz镜像包上传到hd2.com和hd3.com上，并将其上传为镜像：docker load -i nginx.tar.gz

再进入到图形化界面：点击右上角+   create from form

创建新的pod

依次输入信息：App name -> Container image -> Number of pods ->Server(选取External) -> Port-80->Deploy

完成添加

在左侧选择service可以看到端口，在宿主机访问主机IP和该端口，可以访问到nginx的测试页面，即为成功。

安装metrics-server组件

metrics-server是一个集群范围内的资源数据集和工具，同样的metrics-server也只是显示数据，并不提供数据存储服务，主要关注的是资源度量api的实现，比如cpu，文件描述，内存，请求延时等指标，metric-server收集数据给k8s集群内使用，如kubectl，hpa，scheduler等，

部署metrics-server组件。

将metrics-server-amd64-0-3-6.tar.gz，addon.tar.gz上传至镜像仓库

三台主机都进行操作：docker load -i metrics-server-amd64-0-3-6.tar.gz

                                     docker load -i addon.tar.gz

部署metrics-server服务

在/etc/kubernetes/manifests里面改一下apiserver的配置

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

增加如下内容：

在command模块的第四行下添加：---enable-aggregator-routing=true

重新更新apiserver配置：

kubectl apply -f /etc/kubectl/manifests/kube-apiserver.yaml

kubectl get pods - n kube-system

会有一个拉取失败的将失败的进行删除：

kubectl delete pods kube-apiserver -n kube-system

上传metrics.yaml文件：

vim metrics.yaml由于文件较长，在文件夹中k8s使用压缩包中存放，可以自取

kubectl apply -f metrics.yaml

kubectl get pods -n kube-system| grep metrics

出现运行后即可进行下一步。

测试kybectl top命令

kubectl top pods -n kube-system

kubectl top nodes

将scheduler，confroller-manager端口变成物理机可以监听的端口

kubectl get cs

默认在1.19之后10252和10251都是绑定在127的如果想要通过prometheus监控，会采集不到数据，所以可以把端口绑定到物理机按照下列方法进行：

vim /etc/kubernetes/manifests/kube-scheduler.yaml

将--bind-address=127.0.0.1变成--bind-address=192.168.1.11

-port=0删除掉

将httpGet字段下的hosts由127.0.0.1变成192.168.1.11（192.168.1.11是k8s的控制节点的IP）

vim /etc/kubernetes/manifests/kube-confroller-manager.yaml