目录
- 前言
- 一、表象
- 二、解决方法
- 1、导入镜像包
- 2、编辑yaml文件
- 3、解决问题
- 三、优化改造
- 1.修改配置文件
- 2.检查api-server服务是否正常
- 3.测试验证
- 总结
前言
各位老铁大家好，好久不见，卑微涛目前从事kubernetes相关容器工作，感兴趣的小伙伴相互交流一下鸭~
一、表象
使用kubeadm、二进制方式安装的K8S，想查看集群中，node节点/pod的CUP、内存等信息，无法查看
报错：Error from server (ServiceUnavailable): the server is currently unable to handle the request (get pods.metrics.k8s.io)
原因是：缺少metrics-server这个pod
二、解决方法
下面咱们手把手解决这个问题
1、导入镜像包
导入下列两个镜像包
镜像包链接：【若链接失效请私聊卑微涛】
链接：https://pan.baidu.com/s/1qo6QTqF9xSEfeN9OS9BjdQ
提取码：gjx4
#如果k8s是 docker 作为 容器运行时
docker image load -i addon.tar.gz
docker image load -i metrics-server-amd64-0-3-6.tar.gz#如果k8s是 container 作为 容器运行时
ctr -n=k8s.io images import addon.tar.gz
ctr -n=k8s.io images import metrics-server-amd64-0-3-6.tar.gz
#==================================================================
#查看k8s是哪个作为 容器运行时
kubectl get node -o wide
2、编辑yaml文件
注意修改对应的 images：xxx 镜像名称，其余配置无需修改
# cat metrics.yaml #这个yaml文件在镜像包对应的目录中有，大家下载也行
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:name: metrics-server:system:auth-delegatorlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
roleRef:apiGroup: rbac.authorization.k8s.iokind: ClusterRolename: system:auth-delegator
subjects:
- kind: ServiceAccountname: metrics-servernamespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:name: metrics-server-auth-readernamespace: kube-systemlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
roleRef:apiGroup: rbac.authorization.k8s.iokind: Rolename: extension-apiserver-authentication-reader
subjects:
- kind: ServiceAccountname: metrics-servernamespace: kube-system
---
apiVersion: v1
kind: ServiceAccount
metadata:name: metrics-servernamespace: kube-systemlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:name: system:metrics-serverlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
rules:
- apiGroups:- ""resources:- pods- nodes- nodes/stats- namespacesverbs:- get- list- watch
- apiGroups:- "extensions"resources:- deploymentsverbs:- get- list- update- watch
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:name: system:metrics-serverlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
roleRef:apiGroup: rbac.authorization.k8s.iokind: ClusterRolename: system:metrics-server
subjects:
- kind: ServiceAccountname: metrics-servernamespace: kube-system
---
apiVersion: v1
kind: ConfigMap
metadata:name: metrics-server-confignamespace: kube-systemlabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: EnsureExists
data:NannyConfiguration: |-apiVersion: nannyconfig/v1alpha1kind: NannyConfiguration
---
apiVersion: apps/v1
kind: Deployment
metadata:name: metrics-servernamespace: kube-systemlabels:k8s-app: metrics-serverkubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcileversion: v0.3.6
spec:selector:matchLabels:k8s-app: metrics-serverversion: v0.3.6template:metadata:name: metrics-serverlabels:k8s-app: metrics-serverversion: v0.3.6spec:priorityClassName: system-cluster-criticalserviceAccountName: metrics-servercontainers:- name: metrics-serverimage: k8s.gcr.io/metrics-server-amd64:v0.3.6 #这里的镜像名称，根据导入的镜像包修改imagePullPolicy: IfNotPresentcommand:- /metrics-server- --metric-resolution=30s- --kubelet-preferred-address-types=InternalIP- --kubelet-insecure-tlsports:- containerPort: 443name: httpsprotocol: TCP- name: metrics-server-nannyimage: k8s.gcr.io/addon-resizer:1.8.4 #这里的镜像名称，根据导入的镜像包修改imagePullPolicy: IfNotPresentresources:limits:cpu: 100mmemory: 300Mirequests:cpu: 5mmemory: 50Mienv:- name: MY_POD_NAMEvalueFrom:fieldRef:fieldPath: metadata.name- name: MY_POD_NAMESPACEvalueFrom:fieldRef:fieldPath: metadata.namespacevolumeMounts:- name: metrics-server-config-volumemountPath: /etc/configcommand:- /pod_nanny- --config-dir=/etc/config- --cpu=300m- --extra-cpu=20m- --memory=200Mi- --extra-memory=10Mi- --threshold=5- --deployment=metrics-server- --container=metrics-server- --poll-period=300000- --estimator=exponential- --minClusterSize=2volumes:- name: metrics-server-config-volumeconfigMap:name: metrics-server-confignodeSelector:galaxy-app: kce-monitortolerations:- key: "CriticalAddonsOnly"operator: "Exists"- key: node-role.kubernetes.io/mastereffect: NoSchedule
---
apiVersion: v1
kind: Service
metadata:name: metrics-servernamespace: kube-systemlabels:addonmanager.kubernetes.io/mode: Reconcilekubernetes.io/cluster-service: "true"kubernetes.io/name: "Metrics-server"
spec:selector:k8s-app: metrics-serverports:- port: 443protocol: TCPtargetPort: https
---
apiVersion: apiregistration.k8s.io/v1
kind: APIService
metadata:name: v1beta1.metrics.k8s.iolabels:kubernetes.io/cluster-service: "true"addonmanager.kubernetes.io/mode: Reconcile
spec:service:name: metrics-servernamespace: kube-systemgroup: metrics.k8s.ioversion: v1beta1insecureSkipTLSVerify: truegroupPriorityMinimum: 100versionPriority: 100
kubectl apply -f metrics.yaml #运行这个yaml文件
3、解决问题
查看对应的pod是否正常运行
kubectl get deploy -n kube-system metrics-serverkubectl get pod -n kube-system |grep metrics-server
等待30秒，即可正常使用了
三、优化改造
作为一名"老运维"工程师（好吧，就2年😉）,优化+高可用是我们必须考虑的，有一次演练把这个pod删后，虽然重新拉起了，但kubectl top pod 又异常了，我们需要再做这么一个操作
1.修改配置文件
# 修改每个 API Server 的 kube-apiserver.yaml 配置开启 Aggregator Routing：修改 manifests 配置后 API Server 会自动重启生效。
$ cat /etc/kubernetes/manifests/kube-apiserver.yaml
spec:containers:- command:- --enable-aggregator-routing=true #增加这一行配置
注意：/etc/kubernetes/manifests这个目录下的文件由kubelet服务检测，是静态pod，修改完即可，不需要再kubectl apply -f 运行，会自动重新建立对应的pod
2.检查api-server服务是否正常
$ kubectl describe svc metrics-server -n kube-system# 在其他几个节点ping一下Endpoints的地址
ping IP
telnet IP 443
3.测试验证
我们再测试一下，把metrics-server对应的pod删除后，是否能正常使用 kubectl top 指令
大功告成！
总结
"失踪"的一年多时间里，经历了很多，从数据库 --> 大数据 --> 容器，每一次改变，逼着自己跳出舒适圈，
感谢这一路的挫折，让我变得更Strong！
后续不断输出 容器&Kubernetes 相关博客
我是卑微涛，咱们下一篇文章再见了🤞
