# install-k8s-1.9.1 **Repository Path**: jack2zhang/install-k8s-1.9.1 ## Basic Information - **Project Name**: install-k8s-1.9.1 - **Description**: install-k8s-1.9.1 - **Primary Language**: Unknown - **License**: Not specified - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2019-05-13 - **Last Updated**: 2021-11-02 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README ## install-k8s ### /etc/hosts ``` 192.168.9.171 k8s-master-1 192.168.9.172 k8s-master-2 192.168.9.173 k8s-master-3 192.168.9.174 k8s-node-1 192.168.9.175 k8s-node-2 ``` ### hostname ``` hostnamectl --static set-hostname k8s-master-1 hostnamectl --static set-hostname k8s-master-2 hostnamectl --static set-hostname k8s-master-3 hostnamectl --static set-hostname k8s-node-1 hostnamectl --static set-hostname k8s-node-2 ``` ### install cfssl ### 免密 ssh-keygen -t rsa ssh-keygen -t rsa -f ~/.ssh/id_rsa -N '' ``` ssh-copy-id 192.168.9.171 ssh-copy-id 192.168.9.172 ssh-copy-id 192.168.9.173 ssh-copy-id 192.168.9.174 ssh-copy-id 192.168.9.175 ``` 自动设置免密的脚本 ssh-cpoy ### install cfssl ``` yum install etcd -y ``` - 修改etcd-csr.json里面的IP - 生成etcd的证书，分发到所有节点，包括node节点 ``` ./genCerts.sh ./deliver-pem.sh 192.168.9.171 ./deliver-pem.sh 192.168.9.172 ./deliver-pem.sh 192.168.9.173 ./deliver-pem.sh 192.168.9.174 ./deliver-pem.sh 192.168.9.175 ``` - 分发配置文件并修改 - 启动服务 - 检查服务状态 ``` export ETCDCTL_API=3 etcdctl --cacert=/etc/etcd/ssl/etcd-root-ca.pem --cert=/etc/etcd/ssl/etcd.pem --key=/etc/etcd/ssl/etcd-key.pem --endpoints=https://192.168.18.181:2379,https://192.168.18.182:2379,https://192.168.18.183:2379 endpoint health ``` 具体etcd的安装配置参考install-etcd项目 （install-etcd）https://gitee.com/jack2zhang/install-etcd ### 安装k8s的2进制文件 bin目录 - 下载 ``` wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kube-apiserver wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kube-controller-manager wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kube-scheduler wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kubectl wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kubelet wget http://test.wgmf.com/mirrors/kubernetes/v191/bin/kube-proxy ``` - 分发 ``` deliver-bin-master.sh 192.168.9.171 deliver-bin-master.sh 192.168.9.172 deliver-bin-master.sh 192.168.9.173 deliver-bin-node.sh 192.168.9.174 deliver-bin-node.sh 192.168.9.175 ``` ### 生成和分发证书 master和node的 config配置文件 certs/kubernetes - 生成证书 改kubernetes-csr.json 文件里面的IP master和node 要分别生成 config ``` genCerts.sh 生成证书 genTokenKubeConfig.sh 生成master的config和token set-kubelet-config.sh 生成node的config ``` - 分发证书 ``` ./deliver-pem-master.sh 192.168.9.171 ./deliver-pem-master.sh 192.168.9.172 ./deliver-pem-master.sh 192.168.9.173 ./deliver-pem-node.sh 192.168.9.174 ./deliver-pem-node.sh 192.168.9.175 ``` ### 分发systemd 文件 - 分发完需要改IP ``` ./deliver-systemd-master.sh 192.168.9.171 ./deliver-systemd-master.sh 192.168.9.172 ./deliver-systemd-master.sh 192.168.9.173 ``` - master启动 kube-apiserver.service kube-controller-manager kube-scheduler.service ``` start-systemd-master.sh ``` - master组件检查 ``` kubectl get cs ``` ### node 安装 docker docker/ ``` install_docker4centos7.sh ``` ### node 安装 nginx-proxy nginx.conf改成master的IP 部署 ``` deploy-nginx.sh 192.168.9.174 deploy-nginx.sh 192.168.9.175 ``` ### node 安装 kube组件 master上执行授权 创建请求证书 ``` kubectl create clusterrolebinding kubelet-bootstrap --clusterrole=system:node-bootstrapper --user=kubelet-bootstrap ``` - 分发node的systemd # 此处重复了 ``` ./deliver-systemd-node.sh 192.168.18.174 ./deliver-systemd-node.sh 192.168.18.175 ``` 改IP 和 hostname 并启动 ``` /deliver-systemd-node.sh 192.168.18.174 /deliver-systemd-node.sh 192.168.18.175 ``` systemd路径 /etc/systemd/system/ 改IP和hostname 下载基础镜像 docker-wrapper ./docker-wrapper.py pull gcr.io/google_containers/pause-amd64:3.0 - 安装ipvs ``` yum install ipvsadm -y ``` 启动 ``` systemctl start kubelet.service systemctl enable kubelet.service systemctl start kube-proxy.service systemctl enable kube-proxy.service ``` - 限制node的调度 ``` kubectl cordon k8s-node-1 kubectl uncordon k8s-node-1 ``` - master配置认证 kubectl get csr ``` kubectl get csr | grep Pending | awk '{print $1}' | xargs kubectl certificate approve ``` kubectl get nodes ### Flannel docker conf ``` mv /usr/lib/systemd/system/docker.service.d/flannel.conf /etc/systemd/system/docker.service.d ``` set-etcdctl-flannel ``` etcdctl --endpoints=https://192.168.18.171:2379,https://192.168.18.172:2379,https://192.168.18.173:2379 --cert-file=/etc/etcd/ssl/etcd.pem --ca-file=/etc/etcd/ssl/etcd-root-ca.pem --key-file=/etc/etcd/ssl/etcd-key.pem set /flannel/network/config \ '{"Network":"10.254.64.0/18","SubnetLen":24,"Backend":{"Type":"host-gw"}}' ``` /etc/sysconfig/flanneld ``` FLANNEL_ETCD_ENDPOINTS="https://192.168.18.171:2379,https://192.168.18.172:2379,https://192.168.18.173:2379" FLANNEL_ETCD_PREFIX="/flannel/network" FLANNEL_OPTIONS="-ip-masq=true -etcd-cafile=/etc/etcd/ssl/etcd-root-ca.pem -etcd-certfile=/etc/etcd/ssl/etcd.pem -etcd-keyfile=/etc/etcd/ssl/etcd-key.pem -iface=eth0" ``` 重新设置docker docker/ set-docker-kubernetes.sh ### 安装helm * node ``` yum install socat -y docker-wrapp️er.py pull gcr.io/kubernetes-helm/tiller:v2.14.0 ``` * master install-helm.sh ### coredns - install coredns coredns/ exam.sh ./deploy.sh -r 10.254.0.0/18 -i 10.254.0.2 | kubectl apply -f - - 查看coredns的POD ``` kubectl get pods -n kube-system ``` - test coredns testcore/ kubectl apply -f alpine.yaml test ``` kubectl get pods,svc kubectl exec -it alpine nslookup nginx-svc kubectl exec -it alpine nslookup kubernetes ``` ### ISSUE - 生成和同步etcd.conf的配置