kubeadm部署metrics-server

需求：生产环境是kubeadm部署的v1.22.2版本的k8s，统计资源时发现这套环境没有部署metrics-server这个服务，今天来部署一下

1、在github社区找到这个项目并下载

root@jumpserver-cmcc:~# wget https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

#github下载下来的yaml里面的镜像地址是谷歌的，需要替换成阿里的，这里我用的是我自己的阿里云镜像仓库地址
root@jumpserver-cmcc:~# sed -i s#registry.k8s.io/metrics-server/metrics-server:v0.7.1#registry.cn-hangzhou.aliyuncs.com/qinge/metrics-server:v0.6.0#g  components.yaml
root@jumpserver-cmcc:~# cat components.yaml |grep image
        image: registry.cn-hangzhou.aliyuncs.com/qinge/metrics-server:v0.6.0
        imagePullPolicy: IfNotPresent

2、部署服务

root@jumpserver-cmcc:~# kubectl apply -f components.yaml
root@master01:~# kubectl get pod -n kube-system
NAME                                      READY   STATUS    RESTARTS        AGE
calico-kube-controllers-56d97d6bc-5vfwf   1/1     Running   23 (25m ago)    12d
calico-node-5rtg9                         1/1     Running   15 (25m ago)    12d
calico-node-c6bkp                         1/1     Running   19 (25m ago)    12d
calico-node-fprwx                         1/1     Running   11 (25m ago)    12d
calico-node-qqpkl                         1/1     Running   18 (3h9m ago)   12d
calico-typha-c6d6b4cf7-vbtnn              1/1     Running   19 (3h9m ago)   12d
coredns-7b5944fdcf-rgq5p                  1/1     Running   13 (25m ago)    12d
coredns-7b5944fdcf-ztkdc                  1/1     Running   13 (25m ago)    12d
etcd-master01                             1/1     Running   21 (3h9m ago)   12d
kube-apiserver-master01                   1/1     Running   21 (3h9m ago)   12d
kube-controller-manager-master01          1/1     Running   21 (3h9m ago)   12d
kube-proxy-67x4b                          1/1     Running   19 (3h9m ago)   12d
kube-proxy-bn4wj                          1/1     Running   15 (25m ago)    12d
kube-proxy-slxx7                          1/1     Running   15 (25m ago)    12d
kube-proxy-t2r58                          1/1     Running   11 (25m ago)    12d
kube-scheduler-master01                   1/1     Running   21 (3h9m ago)   12d
kube-state-metrics-5c45cbb87d-wbczr       0/1     Running   8 (25s ago)     4m2s

3、pod中的服务一直起不来

3.1查看pod起不来的原因

root@jumpserver-cmcc:~# kubectl logs -f metrics-server-b798ffd7f-7pjx9 -n kube-system

报错如下：很明显是证书有问题

3.2解决方案

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    k8s-app: metrics-server
  name: metrics-server
  namespace: kube-system
spec:
  selector:
    matchLabels:
      k8s-app: metrics-server
  strategy:
    rollingUpdate:
      maxUnavailable: 0
  template:
    metadata:
      labels:
        k8s-app: metrics-server
    spec:
      containers:
      - args:
        - --cert-dir=/tmp
        - --secure-port=10250           #将端口改成10250
        - --kubelet-insecure-tls        #添加这行，跳过证书的认证
        - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
        - --kubelet-use-node-status-port
        - --metric-resolution=15s

#在重新部署一下，服务起来了
 root@jumpserver-cmcc:~# kubectl get pod -n kube-system
NAME                                    READY   STATUS    RESTARTS       AGE
cilium-6fp4z                            1/1     Running   0              350d
cilium-cjtp5                            1/1     Running   0              350d
cilium-fkmhk                            1/1     Running   0              350d
cilium-g7fvc                            1/1     Running   0              350d
cilium-lztjd                            1/1     Running   0              137d
cilium-operator-69c994665d-rhbvg        1/1     Running   1 (350d ago)   350d
cilium-t4thg                            1/1     Running   0              350d
cilium-tvfq4                            1/1     Running   0              137d
coredns-7bdbbf6bf5-qcm5g                1/1     Running   0              350d
coredns-7bdbbf6bf5-v2khw                1/1     Running   0              350d
etcd-kcs01master01                      1/1     Running   2              350d
etcd-kcs01master02                      1/1     Running   0              350d
etcd-kcs01master03                      1/1     Running   0              350d
kube-apiserver-kcs01master01            1/1     Running   2              350d
kube-apiserver-kcs01master02            1/1     Running   0              350d
kube-apiserver-kcs01master03            1/1     Running   1 (350d ago)   350d
kube-controller-manager-kcs01master01   1/1     Running   3 (350d ago)   350d
kube-controller-manager-kcs01master02   1/1     Running   0              350d
kube-controller-manager-kcs01master03   1/1     Running   0              350d
kube-proxy-5hbqp                        1/1     Running   0              350d
kube-proxy-dqg2l                        1/1     Running   0              137d
kube-proxy-pqmnf                        1/1     Running   0              350d
kube-proxy-sg8rr                        1/1     Running   0              350d
kube-proxy-vghd7                        1/1     Running   0              137d
kube-proxy-w46pw                        1/1     Running   0              350d
kube-proxy-w977k                        1/1     Running   0              350d
kube-scheduler-kcs01master01            1/1     Running   3 (350d ago)   350d
kube-scheduler-kcs01master02            1/1     Running   0              350d
kube-scheduler-kcs01master03            1/1     Running   0              350d
metrics-server-b798ffd7f-7pjx9          1/1     Running   0              21m



#如果是二进制部署的在kube-apiserver的启动文件添加这参数--kubelet-insecure-tls
# vim /etc/systemd/system/kube-apiserver.service
[Service]
Environment="KUBE_API_ARGS=--kubelet-insecure-tls"