Kubernetes 高可用集群需要实现：

• API server 组件高可用
• etcd 集群高可用，需要至少3台服务器

所以 Kubernetes 高可用集群需要至少3台 master，2台 node。

安装前检查项：

• 服务器最小要求硬件配置 CPU 2C，内存 2GB
• 每个服务器hostname，MAC，product_uuid 必须唯一

  # 查看 product_uuid
  cat /sys/class/dmi/id/product_uuid
  

• 修改 SELINUX 配置

  setenforce 0
  sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config
  

• 配置 iptable/ipvs 流量转发功能

  cat <<EOF >  /etc/sysctl.d/k8s.conf
  net.bridge.bridge-nf-call-ip6tables = 1
  net.bridge.bridge-nf-call-iptables = 1
  net.ipv4.ip_forward = 1
  EOF
  sysctl --system
  # 检查 iptable 是否加载
  lsmod | grep br_netfilter
  

• 关闭Swap。参考 CentOS 关闭 swap
• 以下端口不能被其他程序占用

安装 IPVS

# 每次启动都加载 ipvs 所需模块
cat <<EOF > /etc/sysconfig/modules/ipvs.modules 
#!/bin/bash
modprobe -- ip_vs
modprobe -- ip_vs_rr
modprobe -- ip_vs_wrr
modprobe -- ip_vs_sh
modprobe -- nf_conntrack_ipv4
EOF
chmod 755 /etc/sysconfig/modules/ipvs.modules 
bash /etc/sysconfig/modules/ipvs.modules 
# 检查 ipvs 是否加载
lsmod | grep -e ip_vs -e nf_conntrack_ipv4
# 安装 ipvs 管理工具
yum install -y ipset
yum install -y ipvsadm
touch /etc/sysconfig/ipvsadm
systemctl start ipvsadm
systemctl enable ipvsadm

安装 Docker

参考 CentOS 安装Docker

安装 Keepalive + HAProxy

为了实现 Kubernetes 集群高可用，需要使用 Keepalive + HAProxy 方式对 API server 进行反向代理，通过 VIP 方式访问 API server。
参考 CentOS 安装 Keepalive
参考 CentOS 安装 HAProxy

安装 kubeadm、kubelet、kubectl

1. 添加 Kubernetes 国内源

   cat <<EOF > /etc/yum.repos.d/kubernetes.repo 
   [kubernetes-aliyun]
   name=Kubernetes-aliyun
   baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
   enabled=1
   gpgcheck=1
   repo_gpgcheck=1
   gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg https://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
   
   [kubernetes]
   name=Kubernetes
   baseurl=https://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64
   enabled=0
   gpgcheck=1
   repo_gpgcheck=1
   gpgkey=https://packages.cloud.google.com/yum/doc/yum-key.gpg https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
   EOF
   

2. 安装 kubeadm、kubelet、kubectl

   yum install -y kubelet kubeadm kubectl
   systemctl enable kubelet
   systemctl restart kubelet
   

初始化 Kubernetes 集群

1. 准备 kubeadm-init.yaml 文件

   apiVersion: kubeadm.k8s.io/v1beta2
   kind: ClusterConfiguration
   kubernetesVersion: v1.15.1
   controlPlaneEndpoint: 192.168.1.244:26443    # API server VIP 访问地址
   networking:
     dnsDomain: cluster.local                   # 集群域名后缀
     serviceSubnet: 10.243.0.0/16               # service ip 范围
     podSubnet: 10.244.0.0/16                   # pod ip 范围
   ---
   apiVersion: kubeproxy.config.k8s.io/v1alpha1
   kind: KubeProxyConfiguration
   mode: ipvs                                   # 使用 ipvs 作为流量转发组件，默认为 iptable 
   

2. 在每台节点上拉取安装镜像

   # 从阿里云拉取镜像
   kubeadm config images pull --image-repository registry.aliyuncs.com/google_containers --kubernetes-version v1.15.1
   # 重新tag镜像
   docker tag registry.aliyuncs.com/google_containers/kube-apiserver:v1.15.1 k8s.gcr.io/kube-apiserver:v1.15.1
   docker tag registry.aliyuncs.com/google_containers/kube-controller-manager:v1.15.1 k8s.gcr.io/kube-controller-manager:v1.15.1
   docker tag registry.aliyuncs.com/google_containers/kube-scheduler:v1.15.1 k8s.gcr.io/kube-scheduler:v1.15.1
   docker tag registry.aliyuncs.com/google_containers/kube-proxy:v1.15.1 k8s.gcr.io/kube-proxy:v1.15.1
   docker tag registry.aliyuncs.com/google_containers/pause:3.1 k8s.gcr.io/pause:3.1
   docker tag registry.aliyuncs.com/google_containers/etcd:3.3.10 k8s.gcr.io/etcd:3.3.10
   docker tag registry.aliyuncs.com/google_containers/coredns:1.3.1 k8s.gcr.io/coredns:1.3.1
   
   docker rmi registry.aliyuncs.com/google_containers/kube-apiserver:v1.15.1
   docker rmi registry.aliyuncs.com/google_containers/kube-controller-manager:v1.15.1
   docker rmi registry.aliyuncs.com/google_containers/kube-scheduler:v1.15.1
   docker rmi registry.aliyuncs.com/google_containers/kube-proxy:v1.15.1
   docker rmi registry.aliyuncs.com/google_containers/pause:3.1
   docker rmi registry.aliyuncs.com/google_containers/etcd:3.3.10
   docker rmi registry.aliyuncs.com/google_containers/coredns:1.3.1
   

3. 初始化第一台 master 节点

   kubeadm init --config kubeadm-init.yaml --upload-certs
   # 初始化成功后，console 显示成功信息
   Your Kubernetes control-plane has initialized successfully!
     kubeadm join 192.168.1.244:26443 --token 32y87c.ux6hmk9tkjqntoz3 \
       --discovery-token-ca-cert-hash sha256:bc91781ae11a22b773ff918da00d78685f5548a4891714c276b6a28aac084ce0 \
       --control-plane --certificate-key c1e22a571bb9a671222b3175f5969694670d152318a890c7c8b07231cb2cf4cb
   ......
     kubeadm join 192.168.1.244:26443 --token 32y87c.ux6hmk9tkjqntoz3 \
       --discovery-token-ca-cert-hash sha256:bc91781ae11a22b773ff918da00d78685f5548a4891714c276b6a28aac084ce0
   

4. 复制 kubectl 配置文件到用户目录下

   mkdir -p $HOME/.kube
   sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
   sudo chown $(id -u):$(id -g) $HOME/.kube/config
   

5. 接着安装 flannel 网络组件

   kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
   

6. 检查 pods 是否都正常启动

   kubectl get pods --all-namespaces -o wide
   NAMESPACE     NAME                               READY   STATUS    RESTARTS   AGE    IP              NODE       NOMINATED NODE   READINESS GATES
   kube-system   coredns-5c98db65d4-59hc6           1/1     Running   0          122m   10.244.0.6      master01   <none>           <none>
   kube-system   coredns-5c98db65d4-n7nf6           1/1     Running   0          122m   10.244.0.7      master01   <none>           <none>
   kube-system   etcd-master01                      1/1     Running   0          121m   192.168.1.108   master01   <none>           <none>
   kube-system   kube-apiserver-master01            1/1     Running   0          122m   192.168.1.108   master01   <none>           <none>
   kube-system   kube-controller-manager-master01   1/1     Running   0          122m   192.168.1.108   master01   <none>           <none>
   kube-system   kube-flannel-ds-amd64-tzf78        1/1     Running   0          121m   192.168.1.108   master01   <none>           <none>
   kube-system   kube-proxy-5b6xr                   1/1     Running   0          122m   192.168.1.108   master01   <none>           <none>
   kube-system   kube-scheduler-master01            1/1     Running   0          121m   192.168.1.108   master01   <none>           <none>
   

7. 添加其他 master 节点

   # 在其他 master 节点上使用第一台 master 生成的 token 加入集群
   kubeadm join 192.168.1.244:26443 --token 32y87c.ux6hmk9tkjqntoz3 \
       --discovery-token-ca-cert-hash sha256:bc91781ae11a22b773ff918da00d78685f5548a4891714c276b6a28aac084ce0 \
       --control-plane --certificate-key c1e22a571bb9a671222b3175f5969694670d152318a890c7c8b07231cb2cf4cb
   # 配置 kubectl
   mkdir -p $HOME/.kube
   sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
   sudo chown $(id -u):$(id -g) $HOME/.kube/config
   

8. 添加 node 节点

   # 在 node 节点上使用第一台 master 生成的 token 加入集群
   kubeadm join 192.168.1.244:26443 --token 32y87c.ux6hmk9tkjqntoz3 \
       --discovery-token-ca-cert-hash sha256:bc91781ae11a22b773ff918da00d78685f5548a4891714c276b6a28aac084ce0
   

重置节点

1. 从集群删除节点

   kubectl drain <node name> --delete-local-data --force --ignore-daemonsets
   kubectl delete node <node name>
   

2. 在节点上删除 Kubernetes 组件，重置 iptable/ipvs

   kubeadm reset -f
   iptables -F && iptables -t nat -F && iptables -t mangle -F && iptables -X
   ipvsadm --clear
   rm -rf ~/.kube/