# k8s单节点升级为高可用

参考网址：[https://zahui.fan/posts/34d8fad0/](https://zahui.fan/posts/34d8fad0/)

- 导出kubeadm配置
```
kubectl -n kube-system get configmap kubeadm-config -o jsonpath='{.data.ClusterConfiguration}' > kubeadm.yaml
```
- 添加证书SANs信息

**certSANs到extraArgs中间的内容，既负载均衡地址，所有的master主机名，IP地址**

**controlPlaneEndpoint: kube-api-server:6443   修改成负载均衡的地址**

**如果没有则增加**



```
apiServer:
  certSANs:
    # 这里需要包含负载均衡、所有master节点的hostname和ip
    - kube-api-server
    - 5-10.vpclub.io
    - 5-11.vpclub.io
    - 5-12.vpclub.io
    - 192.168.5.10
    - 192.168.5.11
    - 192.168.5.12
    - 192.168.5.200
  extraArgs:
    authorization-mode: Node,RBAC
  timeoutForControlPlane: 4m0s
apiVersion: kubeadm.k8s.io/v1beta2
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
controlPlaneEndpoint: kube-api-server:6443  # 修改成负载均衡的地址
controllerManager: {}
dns:
  type: CoreDNS
etcd:
  local:
    dataDir: /var/lib/etcd
imageRepository: registry.aliyuncs.com/google_containers
kind: ClusterConfiguration
kubernetesVersion: v1.20.15
networking:
  dnsDomain: cluster.local
  podSubnet: 10.244.0.0/16
  serviceSubnet: 10.96.0.0/12
scheduler: {}



```
[![](http://qq829.cn/book/uploads/images/gallery/2023-08/scaled-1680-/h1x4Rkd5vBO19VDb-image-1691428059058.png)](http://qq829.cn/book/uploads/images/gallery/2023-08/h1x4Rkd5vBO19VDb-image-1691428059058.png)



- 备份原kubernetes配置文件
```
mkdir -p /data/vpclub/kubernetes-bak/kubernetes-20240619
cp /etc/kubernetes/** /data/vpclub/kubernetes-bak/kubernetes-20240619 -rf

```

- 生成新的证书

```
# 删除旧的证书
rm /etc/kubernetes/pki/apiserver.key -rf
rm /etc/kubernetes/pki/apiserver.crt

# 生成新的配置
kubeadm init phase certs apiserver --config kubeadm.yaml



# 查看证书内容，应该要多出刚增加的那些主机和IP地址
openssl x509 -in /etc/kubernetes/pki/apiserver.crt -text
```

[![](http://qq829.cn/book/uploads/images/gallery/2023-08/scaled-1680-/iqKnlTINgI5hkig2-image-1691428177862.png)](http://qq829.cn/book/uploads/images/gallery/2023-08/iqKnlTINgI5hkig2-image-1691428177862.png)


- 更新cluster-info配置

```


# server部分修改为负载均衡地址，本次使用的是 kube-api-server
kubectl -n kube-public edit cm cluster-info

```

[![](http://qq829.cn/book/uploads/images/gallery/2023-08/scaled-1680-/F3rEo9b4zXVYjuTu-image-1691428504407.png)](http://qq829.cn/book/uploads/images/gallery/2023-08/F3rEo9b4zXVYjuTu-image-1691428504407.png)


- 将配置更新到集群

```

kubeadm init phase upload-config kubeadm --config kubeadm.yaml


# 再次查看配置是否已经生效，如果有不正确的地方需要修改过来
kubectl edit cm kubeadm-config -n kube-system



```


- 重启Apiserver


```
kubectl delete pod 	kube-apiserver-5-10.vpclub.io kube-controller-manager-5-10.vpclub.io -n kube-system
```



- 创建KUBE-VIP自动部署清单

```

# 注意命令行中的VIP地址，网卡名称

docker run --network host --rm swr.cn-south-1.myhuaweicloud.com/vp-whdev/all-in-devops/kube-vip:v0.6.0 manifest pod --interface=eth0 --vip 192.168.5.200  --controlplane --services  --arp  --leaderElection | tee /etc/kubernetes/manifests/kube-vip.yaml

# 测试一下VIP是否起效,到其他机器也ping一下
ping kube-api-server
ping 192.168.5.200


```

- 更新其他配置,将所有如下三个文件的server部分修改为负载均衡地址，本次使用的是 kube-api-server

```
vi /etc/kubernetes/kubelet.conf
# 不确定，默认情况下应该为主机IP
vi /etc/kubernetes/controller-manager.conf
# 不确定，默认情况下应该为主机IP
vi /etc/kubernetes/scheduler.conf
```

[![](http://qq829.cn/book/uploads/images/gallery/2023-08/scaled-1680-/JJqvZj4ChFO7sKen-image-1691428541769.png)](http://qq829.cn/book/uploads/images/gallery/2023-08/JJqvZj4ChFO7sKen-image-1691428541769.png)






- 重启kubelet和容器

```
systemctl restart kubelet
kubectl delete pod -n kube-system kube-controller-manager-5-10.vpclub.io
kubectl delete pod -n kube-system kube-scheduler-5-10.vpclub.io
```

- 修改kube-proxy配置，将server 部分修改为负载均衡地址，本次使用的是 kube-api-server
```
kubectl edit configmap kube-proxy -n kube-system

# 重启 kube-proxy 
kubectl rollout restart daemonset kube-proxy -n kube-system
```

- 修改kubectl 配置
```
vi ~/.kube/config
vi /etc/kubernetes/admin.conf
# 查看集群信息
kubectl cluster-info
```

- 加入新master集群

```
echo "$(kubeadm token create --print-join-command) --control-plane --certificate-key $(kubeadm init phase upload-certs --upload-certs | tail -1)"

复制回显到新的master机器执行


```
- 加入node到集群
```
kubeadm token create --print-join-command
```