# metrics-server

**Repository Path**: shinstein/metrics-server

## Basic Information

- **Project Name**: metrics-server
- **Description**: k8s集群之metrics-server
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2023-10-17
- **Last Updated**: 2023-10-17

## Categories & Tags

**Categories**: Uncategorized

**Tags**: Kubernetes, metrics-server

## README

# metrics-server

#### 介绍
k8s集群之metrics-server, 本库下 metrics-server 目录为 metrics-server 的chart 包。
安装笔记：私有-https://www.notion.so/metrics-server-d010b671366a4df5a77cdeb8fb9593b8 公共- https://hushed-sardine-f46.notion.site/metrics-server-d010b671366a4df5a77cdeb8fb9593b8?pvs=4

#### 软件架构
k8s集群2.5.4安装好后，执行 kubectl top no 提示 error: Metrics API not available ，原因是 k8s 1.8版本之后  metrics-server 需要单独安装


#### 安装教程

1.  下载helm包：https://github.com/kubernetes-sigs/metrics-server/releases页面下载 https://github.com/kubernetes-sigs/metrics-server/releases/download/metrics-server-helm-chart-3.11.0/metrics-server-3.11.0.tgz
2.  在values.yaml中配置 

镜像源：
```
image:
  repository: docker.io/bitnami/metrics-server
  # Overrides the image tag whose default is v{{ .Chart.AppVersion }}
  tag: "0.6.4"
```

非tls
```
defaultArgs:
  - --cert-dir=/tmp
  - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
  - --kubelet-use-node-status-port
  - --metric-resolution=15s
  - --kubelet-insecure-tls=true
  - --kubelet-preferred-address-types=InternalIP
```

增加  `- --kubelet-insecure-tls=true` 和 `- --kubelet-preferred-address-types=InternalIP`，否则启动会报错：
```
[root@master metrics-server]# kubectl -n kube-system describe po metrics-server-64b4b85477-29k8s
...
Events:
  Type     Reason     Age               From               Message
  ----     ------     ----              ----               -------
  Normal   Scheduled  27s               default-scheduler  Successfully assigned kube-system/metrics-server-64b4b85477-29k8s to node1
  Normal   Pulled     27s               kubelet            Container image "docker.io/bitnami/metrics-server:0.6.4-debian-11-r70" already present on machine
  Normal   Created    26s               kubelet            Created container metrics-server
  Normal   Started    26s               kubelet            Started container metrics-server
  Warning  Unhealthy  26s               kubelet            Readiness probe failed: Get "https://10.244.1.35:8443/readyz": dial tcp 10.244.1.35:8443: connect: connection refused
  Warning  Unhealthy  7s (x3 over 25s)  kubelet            Readiness probe failed: HTTP probe failed with statuscode: 500
[root@master metrics-server]#
```

3.  安装 `helm upgrade -i metrics-server metrics-server`

#### 使用说明

1.  kubectl top no
```
[root@master data]# kubectl top no
NAME     CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%   
master   155m         3%     925Mi           25%       
node1    72m          1%     1187Mi          44%       
node2    63m          1%     1057Mi          39%       
node3    65m          1%     935Mi           34%       
[root@master data]# 
```