# Centos8下K8S master节点更换IP > 作者/来源: UCloud 运营管理员 > 发布时间: 2023-01-11T05:20:00.000Z > 分类: 其他 > 标签: AI, Kubernetes, Docker > 原文链接: http://117.50.162.249:3000/yun/articles/795 --- # Centos8下K8S master节点更换IP > 来源: https://www.ucloud.cn/yun/130092.html > 作者: IT那活儿 > 发布日期: 发布于2023-01-11 13:20 Centos8下K8S master节点更换IP [ 一:背景 ] 网卡由于采用的DHCP动态获取IP,新创建的机器占用了K8Smaster节点的IP地址,K8S启动报错无法访问主节点,所以更换K8S的IP地址 报错如下: ![](https://ucloud-blog.cn-bj.ufileos.com/articles/130092/images/130092_000.png) [ 二:更换步骤 ] ## ## 2.1查看当前IP地址 ![](https://ucloud-blog.cn-bj.ufileos.com/articles/130092/images/130092_001.png) ## 2.2 到K8Smaster节点/etc/kubernetes/manifests目录下,我们主要修改etcd.yaml,kube-apiserver.yaml这两个配置文件 ![](https://ucloud-blog.cn-bj.ufileos.com/articles/130092/images/130092_002.png) ## 2.3 把配置文件etcd.yaml中,替换以下几个地方的IP地址 ![](https://ucloud-blog.cn-bj.ufileos.com/articles/130092/images/130092_003.png) 把配置文件kube-apiserver.yaml中,替换以下几个地方的IP地址 ![](https://ucloud-blog.cn-bj.ufileos.com/articles/130092/images/130092_004.png) ## 2.4 生成新的admin用config文件 | | | --- | | cd /etc/Kubernetes mv admin.conf admin.conf\_bak 使用如下命令生成新的admin.conf # kubeadm init phase kubeconfig admin --apiserver-advertise-address 192.168.37.128 I0920 09:50:30.940689   38240 version.go:252] remote version is much newer: v1.19.2; falling back to: stable-1.18 W0920 09:50:33.261484   38240 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io] [kubeconfig] Writing "admin.conf" kubeconfig file --192.168.37.128 为当前主机的IP地址 | ## 2.5 删除旧的证书,生成新证书 | | | --- | | cd /etc/kubernetes/pki mv apiserver.key apiserver.key.bak mv apiserver.crt apiserver.crt.bak 使用如下命令生成 # kubeadm init phase certs apiserver --apiserver-advertise-address 192.168.37.128 I0920 09:55:31.656513   45703 version.go:252] remote version is much newer: v1.19.2; falling back to: stable-1.18 W0920 09:55:33.787879   45703 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io] [certs] Generating "apiserver" certificate and key [certs] apiserver serving cert is signed for DNS names [k8s1 kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local] and IPs [10.96.0.1 192.168.37.128] --192.168.37.128 为当前主机的IP地址 | ## 2.6 重启docker,kubelet | | | --- | | root@k8s1 ~ # systemctl restart docker root@k8s1 ~ # systemctl restart kubelet | ## ## 2.7 更新kubectl使用的admin.conf配置文件 | | | --- | | cd /etc/kubernetes/ root@k8s1 /etc/kubernetes # kubectl get nodes --kubeconfig=admin.conf NAME   STATUS     ROLES    AGE   VERSION k8s1   Ready      master   126d   v1.18.2 k8s2   NotReady     126d   v1.18.2 k8s3   NotReady     126d   v1.18.2 k8s4   NotReady     126d   v1.18.2 可以看到master节点现在已经起来了 执行mv admin.conf ~/.kube/config,后续可以使用kubectl get nodes查看K8S集群状态了 root@k8s1 ~ # kubectl get nodes NAME   STATUS     ROLES    AGE   VERSION k8s1   Ready      master   126d   v1.18.2 k8s2   NotReady     126d   v1.18.2 k8s3   NotReady     126d   v1.18.2 k8s4   NotReady     126d   v1.18.2 | ## 2.8 Node节点状态问题处理 从2.7可以看到,master已经起来了,但是node节点都是notready的,到k8s2节点查看日志如下,其他node节点k83,k84也是如此 | | | --- | | 使用journalctl命令查看kubelet日志,可以看到K8S2还是连接的原来的IP192.168.37.129 journalctl -f -u kubelet Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.673121  102389 file\_linux.go:60] Unable to read config path "/etc/kubernetes/manifests": path does not exist, ignoring Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.684202  102389 kubelet.go:2267] node "k8s2" not found Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.784485  102389 kubelet.go:2267] node "k8s2" not found Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.873297  102389 kubelet\_node\_status.go:92] Unable to register node "k8s2" with API server: Post https://192.168.37.129:6443/api/v1/nodes: dial tcp 192.168.37.129:6443: connect: connection refused Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.884986  102389 kubelet.go:2267] node "k8s2" not found Sep 20 10:14:58 k8s2 kubelet[102389]: E0920 10:14:58.985714  102389 kubelet.go:2267] node "k8s2" not found Sep 20 10:14:59 k8s2 kubelet[102389]: E0920 10:14:59.021311  102389 controller.go:136] failed to ensure node lease exists, will retry in 3.2s, error: Get https://192.168.37.129:6443/apis/coordination.k8s.io/v1/namespaces/kube-node-lease/leases/k8s2?timeout=10s: dial tcp 192.168.37.129:6443: connect: connection refused Sep 20 10:14:59 k8s2 kubelet[102389]: E0920 10:14:59.073642  102389 reflector.go:178] k8s.io/kubernetes/pkg/kubelet/config/apiserver.go:46: Failed to list \*v1.Pod: Get https://192.168.37.129:6443/api/v1/pods?fieldSelector=spec.nodeName%3Dk8s2&limit=500&resourceVersion=0: dial tcp 192.168.37.129:6443: connect: connection refused | [ 解决办法  ] | | | --- | | 1.从master节点k8s1拷贝ca.crt到node节点对应的目录 scp ca.crt k8s2:/etc/kubernetes/pki/ 2.修改node节点的kubelet.conf,把server: https://192.168.37.129:6443修改为现在master节点的ip地址192.168.37.128 3.重启node节点的docker systemctl restart docker 4.使用kubectl get nodes查看,node状态已经正常 # kubectl get nodes NAME   STATUS   ROLES    AGE   VERSION k8s1   Ready    master   126d   v1.18.2 k8s2   Ready       126d v1.18.2 k8s3   Ready       126d v1.18.2 k8s4   Ready       126d v1.18.2 |