部署Kubernetes v1.22.10高可用集群

一、概述

Kubernetes集群控制平面(Master)节点右数据库服务(Etcd)+其它服务组件（Apiserver、Controller-manager、Scheduler等）组成；整个集群系统运行的交互数据都将存储到数据库服务(Etcd)中，所以Kubernetes集群的高可用性取决于数据库服务(Etcd)在多个控制平面(Master)节点构建的数据同步复制关系。由此搭建Kubernetes的高可用集群可以选择以下两种部署方式：

• 使用堆叠的控制平面（Master）节点，其中etcd与组成控制平面的其他组件在同台机器上；
• 使用外部Etcd节点，其中Etcd与控制平台的其他组件在不同的机器上。

参考文档：https://kubernetes.io/zh-cn/docs/setup/production-environment/tools/kubeadm/high-availability/

1.1 堆叠Etcd拓扑(推荐)

Etcd与其他组件共同运行在多台控制平面（Master）机器上，构建Etcd集群关系以形成高可用的Kubernetes集群。

先决条件：

• 最少三个或更多奇数Master节点；
• 最少三个或更多Node节点；
• 集群中所有机器之间的完整网络连接（公共或专用网络）；
• 使用超级用户权限；
• 在集群中的任何一个节点上都可以使用SSH远程访问；
• Kubeadm和Kubelet已经安装到机器上。

使用这种方案可以减少要使用机器的数量，降低成本，降低部署复杂度；多组件服务之间竞争主机资源，可能导致性能瓶颈，以及当Master主机发生故障时影响到所有组件正常工作。

在实际应用中，你可以选择部署更多数量>3的Master主机，则该拓扑的劣势将会减弱！

这是kubeadm中的默认拓扑，kubeadm会在Master节点上自动创建本地etcd成员。

1.2 外部Etcd拓扑

控制平面的Etcd组件运行在外部主机上，其他组件连接到外部的Etcd集群以形成高可用的Kubernetes集群。

先决条件：

• 最少三个或更多奇数Master主机；
• 最少三个或更多Node主机；
• 还需要三台或更多奇数Etcd主机。
• 集群中所有主机之间的完整网络连接（公共或专用网络）；
• 使用超级用户权限；
• 在集群中的任何一个节点主机上都可以使用SSH远程访问；
• Kubeadm和Kubelet已经安装到机器上。

使用外部主机搭建起来的Etcd集群，拥有更多的主机资源和可扩展性，以及故障影响范围缩小，但更多的机器将导致增加部署成本。

二、部署规划

主机系统：CentOS Linux release 7.7.1908 (Core)

Kubernetes版本：1.22.10

Docker CE版本：20.10.17

管理节点运行服务：etcd、kube-apiserver、kube-scheduler、kube-controller-manager、docker、kubelet、keepalived、haproxy

管理节点配置：4vCPU / 8GB内存 / 200G存储

主机名	主机地址	VIP地址	主机角色
k8s-master01	192.168.0.5	192.168.0.10	Master（Control Plane）
k8s-master02	192.168.0.6		Master（Control Plane）
k8s-master03	192.168.0.7		Master（Control Plane）

注：确保服务器为全新安装的系统，未安装其它软件仅用于Kubernetes运行。
可使用如下命令检查端口是否被占用：

ss -alnupt |grep -E '6443|10250|10259|10257|2379|2380'
ss -alnupt |grep -E '10250|3[0-2][0-7][0-6][0-7]'

三、搭建Kubernetes集群

3.1 内核升级(可选)

CentOS 7.x 版本的系统默认内核是3.10，该版本的内核在Kubernetes社区有很多已知的Bug（如：内核内存泄漏错误），建议升级成4.17+版本以上。

官方镜像仓库下载地址：http://mirrors.coreix.net/elrepo-archive-archive/kernel/el7/x86_64/RPMS/

# 安装4.19.9-1版本内核
$ rpm -ivh http://mirrors.coreix.net/elrepo-archive-archive/kernel/el7/x86_64/RPMS/kernel-ml-4.19.9-1.el7.elrepo.x86_64.rpm
$ rpm -ivh http://mirrors.coreix.net/elrepo-archive-archive/kernel/el7/x86_64/RPMS/kernel-ml-devel-4.19.9-1.el7.elrepo.x86_64.rpm

# 查看内核启动顺序
$ awk -F \' '$1=="menuentry " {print i++ " : " $2}' /etc/grub2.cfg
0 : CentOS Linux (3.10.0-1062.12.1.el7.x86_64) 7 (Core)
1 : CentOS Linux (4.19.9-1.el7.elrepo.x86_64) 7 (Core)
2 : CentOS Linux (3.10.0-862.el7.x86_64) 7 (Core)
3 : CentOS Linux (0-rescue-ef219b153e8049718c374985be33c24e) 7 (Core)

# 设置系统启动默认内核
$ grub2-set-default "CentOS Linux (4.19.9-1.el7.elrepo.x86_64) 7 (Core)"
$ grub2-mkconfig -o /boot/grub2/grub.cfg

# 查看默认内核
$ grub2-editenv list
CentOS Linux (4.19.9-1.el7.elrepo.x86_64) 7 (Core)

# 重启系统使其生效
$ reboot

3.2 系统初始化

3.2.1 设置主机名

### 在master01上执行
$ hostnamectl set-hostname k8s-master01
# 在master02上执行
$ hostnamectl set-hostname k8s-master02
# 在master03上执行
$ hostnamectl set-hostname k8s-master03

3.2.2 添加hosts名称解析

### 在所有主机上执行
$ cat >> /etc/hosts << EOF
192.168.0.5 k8s-master01
192.168.0.6 k8s-master02
192.168.0.7 k8s-master03
EOF

3.2.3 安装常用软件

### 在所有主机上执行
$ yum -y install epel-release.noarch nfs-utils net-tools bridge-utils \
ntpdate vim chrony wget lrzsz

3.2.4 设置主机时间同步

在k8s-master01上设置从公共时间服务器上同步时间

[root@k8s-master01 ~]# systemctl stop ntpd
[root@k8s-master01 ~]# timedatectl set-timezone Asia/Shanghai
[root@k8s-master01 ~]# ntpdate ntp.aliyun.com && /usr/sbin/hwclock
[root@k8s-master01 ~]# vim /etc/ntp.conf
# 当该节点丢失网络连接，采用本地时间作为时间服务器为集群中的其他节点提供时间同步
server 127.127.1.0
Fudge  127.127.1.0 stratum 10
# 注释掉默认时间服务器，改为如下地址
server cn.ntp.org.cn prefer iburst minpoll 4 maxpoll 10
server ntp.aliyun.com iburst minpoll 4 maxpoll 10
server time.ustc.edu.cn iburst minpoll 4 maxpoll 10
server ntp.tuna.tsinghua.edu.cn iburst minpoll 4 maxpoll 10

[root@k8s-master01 ~]# systemctl start ntpd
[root@k8s-master01 ~]# systemctl enable ntpd
[root@k8s-master01 ~]# ntpstat
synchronised to NTP server (203.107.6.88) at stratum 3
   time correct to within 202 ms
   polling server every 64 s

配置其它主机从k8s-master01同步时间

### 在除k8s-master01以外的所有主机上执行
$ systemctl stop ntpd
$ timedatectl set-timezone Asia/Shanghai
$ ntpdate k8s-master01 && /usr/sbin/hwclock
$ vim /etc/ntp.conf
# 注释掉默认时间服务器，改为如下地址
server k8s-master01 prefer iburst minpoll 4 maxpoll 10

$ systemctl start ntpd
$ systemctl enable ntpd
$ ntpstat
synchronised to NTP server (192.168.0.5) at stratum 4
   time correct to within 217 ms
   polling server every 16 s

3.2.5 关闭防火墙

### 在所有节点上执行
# 关闭SElinux
$ sed -i 's/^SELINUX=enforcing$/SELINUX=disabled/' /etc/selinux/config
$ setenforce 0
# 关闭Fileworld防火墙
$ systemctl stop firewalld.service
$ systemctl disable firewalld.service

3.2.6 系统优化

### 在所有节点上执行
# 关闭swap
$ swapoff -a
$ sed -i "s/^[^#].*swap/#&/g" /etc/fstab

# 启用bridge-nf功能
$ cat > /etc/modules-load.d/k8s.conf << EOF
overlay
br_netfilter
EOF
$ modprobe overlay && modprobe br_netfilter

# 设置内核参数
$ cat > /etc/sysctl.d/k8s.conf << EOF
# 配置转发 IPv4 并让 iptables 看到桥接流量
net.ipv4.ip_forward                 = 1
net.bridge.bridge-nf-call-iptables  = 1
net.bridge.bridge-nf-call-ip6tables = 1

# 加强握手队列能力
net.ipv4.tcp_max_syn_backlog        = 10240
net.core.somaxconn                  = 10240
net.ipv4.tcp_syncookies             = 1

# 调整系统级别的能够打开的文件句柄的数量
fs.file-max=1000000

# 配置arp cache 大小
net.ipv4.neigh.default.gc_thresh1   = 1024
net.ipv4.neigh.default.gc_thresh2   = 4096
net.ipv4.neigh.default.gc_thresh3   = 8192

# 令TCP窗口和状态追踪更加宽松
net.netfilter.nf_conntrack_tcp_be_liberal = 1
net.netfilter.nf_conntrack_tcp_loose = 1

# 允许的最大跟踪连接条目，是在内核内存中netfilter可以同时处理的“任务”（连接跟踪条目）
net.netfilter.nf_conntrack_max      = 10485760
net.netfilter.nf_conntrack_tcp_timeout_established = 300
net.netfilter.nf_conntrack_buckets  = 655360

# 每个网络接口接收数据包的速率比内核处理这些包的速率快时，允许送到队列的数据包的最大数目。
net.core.netdev_max_backlog         = 10000

# 默认值: 128 指定了每一个real user ID可创建的inotify instatnces的数量上限
fs.inotify.max_user_instances       = 524288
# 默认值: 8192 指定了每个inotify instance相关联的watches的上限
fs.inotify.max_user_watches         = 524288
EOF
$ sysctl --system 

# 修改文件打开数
$ ulimit -n 65545
$ cat >> /etc/sysctl.d/limits.conf << EOF
*               soft    nproc           65535
*               hard    nproc           65535
*               soft    nofile          65535
*               hard    nofile          65535
EOF
$ sed -i '/nproc/ s/4096/65535/' /etc/security/limits.d/20-nproc.conf

3.3 安装Docker

### 在所有节点上执行
# 安装Docker
$ yum install -y yum-utils device-mapper-persistent-data lvm2
$ yum-config-manager --add-repo https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
$ sed -i 's+download.docker.com+mirrors.aliyun.com/docker-ce+' /etc/yum.repos.d/docker-ce.repo && yum makecache fast
$ yum -y install docker-ce-20.10.17

# 优化docker配置
$ mkdir -p /etc/docker && cat > /etc/docker/daemon.json <<EOF
{
  "registry-mirrors": [
      "https://hub-mirror.c.163.coma",
      "https://docker.mirrors.ustc.edu.cn",
      "https://p6902cz5.mirror.aliyuncs.com"
  ],
  "exec-opts": ["native.cgroupdriver=systemd"],
  "log-driver": "json-file",
  "log-opts": {
    "max-size": "100m"
  },
  "storage-driver": "overlay2",
  "storage-opts": [
    "overlay2.override_kernel_check=true"
  ],
  "bip": "172.38.16.1/24"
}
EOF

# 启动并配置开机自启
$ systemctl enable docker
$ systemctl restart docker
$ docker version

3.4 安装Kubernetes

### 在所有Master节点执行
# 配置yum源
cat > /etc/yum.repos.d/kubernetes.repo <<EOF
[kubernetes]
name=Kubernetes
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
EOF

# 安装kubeadm、kubelet和kubectl
$ yum install -y kubelet-1.22.10 kubeadm-1.22.10 kubectl-1.22.10 --disableexcludes=kubernetes --nogpgcheck
$ systemctl enable --now kubelet

# 配置kubelet参数
$ cat > /etc/sysconfig/kubelet <<EOF
KUBELET_EXTRA_ARGS="--fail-swap-on=false"
EOF

可以参考：https://www.yuque.com/wubolive/ops/ugomse 修改kubeadm源码更改证书签发时长。

3.5 配置HA负载均衡

当存在多个控制平面时，kube-apiserver也存在多个，可以使用HAProxy+Keepalived这个组合，因为HAProxy可以提高更高性能的四层负载均衡功能。

官方文档提供了两种运行方式（此案例使用选项2）：

• 选项1：在操作系统上运行服务
• 选项2：将服务作为静态pod运行

参考文档：https://github.com/kubernetes/kubeadm/blob/main/docs/ha-considerations.md#options-for-software-load-balancing

3.5.1 配置keepalived

将keepalived作为静态pod运行，在引导过程中，kubelet将启动这些进程，以便集群可以在启动时使用它们。这是一个优雅的解决方案，特别是在堆叠（Stacked）etcd 拓扑下描述的设置。

创建keepalived.conf配置文件

### 在k8s-master01上设置：
$ mkdir /etc/keepalived && cat > /etc/keepalived/keepalived.conf <<EOF
! /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
    router_id k8s-master01
}
vrrp_script check_apiserver {
  script "/etc/keepalived/check_apiserver.sh"
  interval 3
  weight -2
  fall 10
  rise 2
}

vrrp_instance VI_1 {
    state MASTER
    interface eth0
    virtual_router_id 51
    priority 100
    authentication {
        auth_type PASS
        auth_pass 123456
    }
    virtual_ipaddress {
        192.168.0.10
    }
    track_script {
        check_apiserver
    }
}
EOF

### 在k8s-master02上设置：
$ mkdir /etc/keepalived && cat > /etc/keepalived/keepalived.conf <<EOF
! /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
    router_id k8s-master02
}
vrrp_script check_apiserver {
  script "/etc/keepalived/check_apiserver.sh"
  interval 3
  weight -2
  fall 10
  rise 2
}

vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 51
    priority 99
    authentication {
        auth_type PASS
        auth_pass 123456
    }
    virtual_ipaddress {
        192.168.0.10
    }
    track_script {
        check_apiserver
    }
}
EOF

### 在k8s-master03上设置：
$ mkdir /etc/keepalived && cat > /etc/keepalived/keepalived.conf <<EOF
! /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
    router_id k8s-master03
}
vrrp_script check_apiserver {
  script "/etc/keepalived/check_apiserver.sh"
  interval 3
  weight -2
  fall 10
  rise 2
}

vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 51
    priority 98
    authentication {
        auth_type PASS
        auth_pass 123456
    }
    virtual_ipaddress {
        192.168.0.10
    }
    track_script {
        check_apiserver
    }
}
EOF

创建健康检查脚本

### 在所有Master控制节点上执行
$ cat > /etc/keepalived/check_apiserver.sh << 'EOF'
#!/bin/sh

errorExit() {
    echo "*** $*" 1>&2
    exit 1
}

curl --silent --max-time 2 --insecure https://localhost:9443/ -o /dev/null || errorExit "Error GET https://localhost:9443/"
if ip addr | grep -q 192.168.0.10; then
    curl --silent --max-time 2 --insecure https://192.168.0.10:9443/ -o /dev/null || errorExit "Error GET https://192.168.0.10:9443/"
fi
EOF

3.5.2 配置haproxy

### 在所有Master管理节点执行
$ mkdir /etc/haproxy && cat > /etc/haproxy/haproxy.cfg << 'EOF'
# /etc/haproxy/haproxy.cfg
#---------------------------------------------------------------------
# Global settings
#---------------------------------------------------------------------
global
    log /dev/log local0
    log /dev/log local1 notice
    daemon

#---------------------------------------------------------------------
# common defaults that all the 'listen' and 'backend' sections will
# use if not designated in their block
#---------------------------------------------------------------------
defaults
    mode                    http
    log                     global
    option                  httplog
    option                  dontlognull
    option http-server-close
    option forwardfor       except 127.0.0.0/8
    option                  redispatch
    retries                 1
    timeout http-request    10s
    timeout queue           20s
    timeout connect         5s
    timeout client          20s
    timeout server          20s
    timeout http-keep-alive 10s
    timeout check           10s

#---------------------------------------------------------------------
# Haproxy Monitoring panel
#---------------------------------------------------------------------
listen  admin_status
    bind 0.0.0.0:8888
    mode http
    log 127.0.0.1 local3 err
    stats refresh 5s
    stats uri /admin?stats
    stats realm itnihao\ welcome
    stats auth admin:admin
    stats hide-version
    stats admin if TRUE

#---------------------------------------------------------------------
# apiserver frontend which proxys to the control plane nodes
#---------------------------------------------------------------------
frontend apiserver
    bind *:9443
    mode tcp
    option tcplog
    default_backend apiserver

#---------------------------------------------------------------------
# round robin balancing for apiserver
#---------------------------------------------------------------------
backend apiserver
    option httpchk GET /healthz
    http-check expect status 200
    mode tcp
    option ssl-hello-chk
    balance     roundrobin
        server k8s-master01 192.168.0.5:6443 check
        server k8s-master02 192.168.0.6:6443 check
        server k8s-master03 192.168.0.7:6443 check
EOF

3.5.3 配置静态Pod运行

对于此设置，需要在其中创建两个清单文件/etc/kubernetes/manifests（首先创建目录）。

### 仅在k8s-master01上创建
$ mkdir -p /etc/kubernetes/manifests
# 配置keepalived清单
$ cat > /etc/kubernetes/manifests/keepalived.yaml << 'EOF'
apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: null
  name: keepalived
  namespace: kube-system
spec:
  containers:
  - image: osixia/keepalived:2.0.17
    name: keepalived
    resources: {}
    securityContext:
      capabilities:
        add:
        - NET_ADMIN
        - NET_BROADCAST
        - NET_RAW
    volumeMounts:
    - mountPath: /usr/local/etc/keepalived/keepalived.conf
      name: config
    - mountPath: /etc/keepalived/check_apiserver.sh
      name: check
  hostNetwork: true
  volumes:
  - hostPath:
      path: /etc/keepalived/keepalived.conf
    name: config
  - hostPath:
      path: /etc/keepalived/check_apiserver.sh
    name: check
status: {}
EOF
# 配置haproxy清单
cat > /etc/kubernetes/manifests/haproxy.yaml << 'EOF'
apiVersion: v1
kind: Pod
metadata:
  name: haproxy
  namespace: kube-system
spec:
  containers:
  - image: haproxy:2.1.4
    name: haproxy
    livenessProbe:
      failureThreshold: 8
      httpGet:
        host: localhost
        path: /healthz
        port: 9443
        scheme: HTTPS
    volumeMounts:
    - mountPath: /usr/local/etc/haproxy/haproxy.cfg
      name: haproxyconf
      readOnly: true
  hostNetwork: true
  volumes:
  - hostPath:
      path: /etc/haproxy/haproxy.cfg
      type: FileOrCreate
    name: haproxyconf
status: {}
EOF

3.6 部署Kubernetes集群

3.6.1 准备镜像

由于国内访问k8s.gcr.io存在某些原因下载不了镜像，所以我们可以在国内的镜像仓库中下载它们（比如使用阿里云镜像仓库。阿里云代理镜像仓库地址：registry.aliyuncs.com/google_containers

### 在所有Master控制节点执行
$ kubeadm config images pull --kubernetes-version=v1.22.10 --image-repository=registry.aliyuncs.com/google_containers

3.6.2 准备ini配置文件

### 在k8s-master01上执行
$ kubeadm config print init-defaults > kubeadm-init.yaml
$ vim kubeadm-init.yaml
apiVersion: kubeadm.k8s.io/v1beta3
bootstrapTokens:
- groups:
  - system:bootstrappers:kubeadm:default-node-token
  token: abcdef.0123456789abcdef
  ttl: 24h0m0s
  usages:
  - signing
  - authentication
kind: InitConfiguration
localAPIEndpoint:
  advertiseAddress: 192.168.0.5
  bindPort: 6443
nodeRegistration:
  criSocket: /var/run/dockershim.sock
  imagePullPolicy: IfNotPresent
  name: k8s-master01
  taints: null
---
controlPlaneEndpoint: "192.168.0.10:9443"
apiServer:
  timeoutForControlPlane: 4m0s
apiVersion: kubeadm.k8s.io/v1beta3
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
controllerManager: {}
dns: {}
etcd:
  local:
    dataDir: /var/lib/etcd
imageRepository: registry.aliyuncs.com/google_containers
kind: ClusterConfiguration
kubernetesVersion: 1.22.10
networking:
  dnsDomain: cluster.local
  serviceSubnet: 10.96.0.0/12
scheduler: {}

配置说明：

• localAPIEndpoint.advertiseAddress：本机apiserver监听的IP地址。
• localAPIEndpoint.bindPort：本机apiserver监听的端口。
• controlPlaneEndpoint：控制平面入口点地址（负载均衡器VIP地址+负载均衡器端口）。
• imageRepository：部署集群时要使用的镜像仓库地址。
• kubernetesVersion：部署集群的kubernetes版本。

3.6.3 初始化控制平面节点

kubeadm在初始化控制平面时会生成部署Kubernetes集群中各个组件所需的相关配置文件在/etc/kubernetes目录下，可以供我们参考。

### 在k8s-master01上执行
# 由于kubeadm命令为源码安装，需要配置一下kubelet服务。
$ kubeadm init phase kubelet-start --config kubeadm-init.yaml
# 初始化kubernetes控制平面
$ kubeadm init --config kubeadm-init.yaml --upload-certs

Your Kubernetes control-plane has initialized successfully!

To start using your cluster, you need to run the following as a regular user:

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

Alternatively, if you are the root user, you can run:

  export KUBECONFIG=/etc/kubernetes/admin.conf

You should now deploy a pod network to the cluster.
Run "kubectl apply -f [podnetwork].yaml" with one of the options listed at:
  https://kubernetes.io/docs/concepts/cluster-administration/addons/

You can now join any number of the control-plane node running the following command on each as root:

  kubeadm join 192.168.0.10:9443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b30e986e80423da7b6b1cbf43ece58598074b2a8b86295517438942e9a47ab0d \
        --control-plane --certificate-key 57360054608fa9978864124f3195bc632454be4968b5ccb577f7bb9111d96597

Please note that the certificate-key gives access to cluster sensitive data, keep it secret!
As a safeguard, uploaded-certs will be deleted in two hours; If necessary, you can use
"kubeadm init phase upload-certs --upload-certs" to reload certs afterward.

Then you can join any number of worker nodes by running the following on each as root:

kubeadm join 192.168.0.10:9443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b30e986e80423da7b6b1cbf43ece58598074b2a8b86295517438942e9a47ab0d

3.6.4 将其它节点加入集群

将控制平面节点加入集群

### 在另外两台Master控制节点执行：
$ kubeadm join 192.168.0.10:9443 --token abcdef.0123456789abcdef \
      --discovery-token-ca-cert-hash sha256:b30e986e80423da7b6b1cbf43ece58598074b2a8b86295517438942e9a47ab0d \
      --control-plane --certificate-key 57360054608fa9978864124f3195bc632454be4968b5ccb577f7bb9111d96597

将工作节点加入集群（可选）

### 如有Node工作节点可使用如下命令
$ kubeadm join 192.168.0.10:9443 --token abcdef.0123456789abcdef \
        --discovery-token-ca-cert-hash sha256:b30e986e80423da7b6b1cbf43ece58598074b2a8b86295517438942e9a47ab0d

将keepalived和haproxy复制到其它Master控制节点

$ scp /etc/kubernetes/manifests/{haproxy.yaml,keepalived.yaml} root@k8s-master02:/etc/kubernetes/manifests/
$ scp /etc/kubernetes/manifests/{haproxy.yaml,keepalived.yaml} root@k8s-master03:/etc/kubernetes/manifests/ 

去掉master污点（可选）

  $ kubectl taint nodes --all node-role.kubernetes.io/master-

3.6.5 验证集群状态

### 可在任意Master控制节点执行
# 配置kubectl认证
$ mkdir -p $HOME/.kube
$ cp -i /etc/kubernetes/admin.conf $HOME/.kube/config

# 查看节点状态
$ kubectl get nodes
NAME           STATUS     ROLES                  AGE     VERSION
k8s-master01   NotReady   control-plane,master   13m     v1.22.10
k8s-master02   NotReady   control-plane,master   3m55s   v1.22.10
k8s-master03   NotReady   control-plane,master   113s    v1.22.10

# 查看pod状态
$ kubectl get pod -n kube-system
NAMESPACE     NAME                                   READY   STATUS    RESTARTS        AGE
kube-system   coredns-7f6cbbb7b8-96hp9               0/1     Pending   0               18m
kube-system   coredns-7f6cbbb7b8-kfmnn               0/1     Pending   0               18m
kube-system   etcd-k8s-master01                      1/1     Running   0               18m
kube-system   etcd-k8s-master02                      1/1     Running   0               9m21s
kube-system   etcd-k8s-master03                      1/1     Running   0               7m18s
kube-system   haproxy-k8s-master01                   1/1     Running   0               18m
kube-system   haproxy-k8s-master02                   1/1     Running   0               3m27s
kube-system   haproxy-k8s-master03                   1/1     Running   0               3m16s
kube-system   keepalived-k8s-master01                1/1     Running   0               18m
kube-system   keepalived-k8s-master02                1/1     Running   0               3m27s
kube-system   keepalived-k8s-master03                1/1     Running   0               3m16s
kube-system   kube-apiserver-k8s-master01            1/1     Running   0               18m
kube-system   kube-apiserver-k8s-master02            1/1     Running   0               9m24s
kube-system   kube-apiserver-k8s-master03            1/1     Running   0               7m23s
kube-system   kube-controller-manager-k8s-master01   1/1     Running   0               18m
kube-system   kube-controller-manager-k8s-master02   1/1     Running   0               9m24s
kube-system   kube-controller-manager-k8s-master03   1/1     Running   0               7m22s
kube-system   kube-proxy-cvdlr                       1/1     Running   0               7m23s
kube-system   kube-proxy-gnl7t                       1/1     Running   0               9m25s
kube-system   kube-proxy-xnrt7                       1/1     Running   0               18m
kube-system   kube-scheduler-k8s-master01            1/1     Running   0               18m
kube-system   kube-scheduler-k8s-master02            1/1     Running   0               9m24s
kube-system   kube-scheduler-k8s-master03            1/1     Running   0               7m22s

# 查看kubernetes证书有效期
$ kubeadm certs check-expiration
CERTIFICATE                EXPIRES                  RESIDUAL TIME   CERTIFICATE AUTHORITY   EXTERNALLY MANAGED
admin.conf                 Oct 25, 2122 07:40 UTC   99y             ca                      no
apiserver                  Oct 25, 2122 07:40 UTC   99y             ca                      no
apiserver-etcd-client      Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
apiserver-kubelet-client   Oct 25, 2122 07:40 UTC   99y             ca                      no
controller-manager.conf    Oct 25, 2122 07:40 UTC   99y             ca                      no
etcd-healthcheck-client    Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
etcd-peer                  Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
etcd-server                Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
front-proxy-client         Oct 25, 2122 07:40 UTC   99y             front-proxy-ca          no
scheduler.conf             Oct 25, 2122 07:40 UTC   99y             ca                      no      

CERTIFICATE AUTHORITY   EXPIRES                  RESIDUAL TIME   EXTERNALLY MANAGED
ca                      Oct 22, 2032 07:40 UTC   99y             no
etcd-ca                 Oct 22, 2032 07:40 UTC   99y             no
front-proxy-ca          Oct 22, 2032 07:40 UTC   99y             no   

查看HAProxy控制台集群状态

访问：http://192.168.0.10:8888/admin?stats 账号密码都为admin

3.6.6 安装CNA插件(calico)

Calico是一个开源的虚拟化网络方案，支持基础的Pod网络通信和网络策略功能。

官方文档：https://projectcalico.docs.tigera.io/getting-started/kubernetes/quickstart

### 在任意Master控制节点执行
# 下载最新版本编排文件
$ kubectl apply -f https://docs.projectcalico.org/manifests/calico.yaml
# 下载指定版本编排文件(可选)
$ curl https://raw.githubusercontent.com/projectcalico/calico/v3.24.0/manifests/calico.yaml -O
# 部署calico
$ kubectl apply -f calico.yaml

# 验证安装
$ kubectl get pod -n kube-system | grep calico
calico-kube-controllers-86c9c65c67-j7pv4   1/1     Running   0               17m
calico-node-8mzpk                          1/1     Running   0               17m
calico-node-tkzs2                          1/1     Running   0               17m
calico-node-xbwvp                          1/1     Running   0               17m

四、集群优化及组件安装

4.1 集群优化

4.1.1 修改NodePort端口范围（可选）

### 在所有Master管理节点执行
$ sed -i '/- --secure-port=6443/a\    - --service-node-port-range=1-32767' /etc/kubernetes/manifests/kube-apiserver.yaml

4.1.2 解决kubectl get cs显示异常问题

### 在所有Master管理节点执行
$ sed -i 's/^[^#].*--port=0$/#&/g' /etc/kubernetes/manifests/{kube-scheduler.yaml,kube-controller-manager.yaml}
# 验证
$ kubectl get cs
Warning: v1 ComponentStatus is deprecated in v1.19+
NAME                 STATUS    MESSAGE                         ERROR
scheduler            Healthy   ok
controller-manager   Healthy   ok
etcd-0               Healthy   {"health":"true","reason":""} 

4.1.3 解决调度器监控不显示问题

### 在所有Master管理节点执行
$ sed -i 's#bind-address=127.0.0.1#bind-address=0.0.0.0#g' /etc/kubernetes/manifests/kube-controller-manager.yaml
$ sed -i 's#bind-address=127.0.0.1#bind-address=0.0.0.0#g' /etc/kubernetes/manifests/kube-scheduler.yaml

4.2 安装Metric-Server

指标服务Metrices-Server是Kubernetes中的一个常用插件，它类似于Top命令，可以查看Kubernetes中Node和Pod的CPU和内存资源使用情况。Metrices-Server每15秒收集一次指标，它在集群中的每个节点中运行，可扩展支持多达5000个节点的集群。

参考文档：https://github.com/kubernetes-sigs/metrics-server

### 在任意Master管理节点执行
$ wget https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml -O metrics-server.yaml
# 修改配置
$ vim metrics-server.yaml
......
    spec:
      containers:
      - args:
        - --cert-dir=/tmp
        - --secure-port=4443
        - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
        - --kubelet-use-node-status-port
        - --metric-resolution=15s
        - --kubelet-insecure-tls  # 不要验证由Kubelets提供的CA或服务证书。
        image: bitnami/metrics-server:0.6.1   # 修改成docker.io镜像
        imagePullPolicy: IfNotPresent
......
# 部署metrics-server
$ kubectl apply -f metrics-server.yaml
# 查看启动状态
$ kubectl get pod -n kube-system -l k8s-app=metrics-server -w
NAME                             READY   STATUS    RESTARTS   AGE
metrics-server-655d65c95-lvb7z   1/1     Running   0          103s
# 查看集群资源状态
$ kubectl top nodes
NAME           CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
k8s-master01   193m         4%     2144Mi          27%
k8s-master02   189m         4%     1858Mi          23%
k8s-master03   268m         6%     1934Mi          24%    

五、附录

5.1 重置节点(危险操作)

当在使用kubeadm init或kubeadm join部署节点出现失败状况时，可以使用以下操作对节点进行重置！

注：重置会将节点恢复到未部署前状态，若集群已正常工作则无需重置，否则将引起不可恢复的集群故障！

$ kubeadm reset -f
$ ipvsadm --clear
$ iptables -F && iptables -X && iptables -Z

5.2 常用查询命令

# 查看Token列表
$ kubeadm token list
TOKEN                     TTL         EXPIRES                USAGES                   DESCRIPTION                                                EXTRA GROUPS
abcdef.0123456789abcdef   22h         2022-10-26T07:43:01Z   authentication,signing   <none>                                                     system:bootstrappers:kubeadm:default-node-token
jgqg88.6mskuadei41o0s2d   40m         2022-10-25T09:43:01Z   <none>                   Proxy for managing TTL for the kubeadm-certs secret        <none>

# 查询节点运行状态
$ kubectl get nodes
NAME           STATUS   ROLES                  AGE   VERSION
k8s-master01   Ready    control-plane,master   81m   v1.22.10
k8s-master02   Ready    control-plane,master   71m   v1.22.10
k8s-master03   Ready    control-plane,master   69m   v1.22.10

# 查看证书到期时间
$ kubeadm certs check-expiration
CERTIFICATE                EXPIRES                  RESIDUAL TIME   CERTIFICATE AUTHORITY   EXTERNALLY MANAGED
admin.conf                 Oct 25, 2122 07:40 UTC   99y             ca                      no
apiserver                  Oct 25, 2122 07:40 UTC   99y             ca                      no
apiserver-etcd-client      Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
apiserver-kubelet-client   Oct 25, 2122 07:40 UTC   99y             ca                      no
controller-manager.conf    Oct 25, 2122 07:40 UTC   99y             ca                      no
etcd-healthcheck-client    Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
etcd-peer                  Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
etcd-server                Oct 25, 2122 07:40 UTC   99y             etcd-ca                 no
front-proxy-client         Oct 25, 2122 07:40 UTC   99y             front-proxy-ca          no
scheduler.conf             Oct 25, 2122 07:40 UTC   99y             ca                      no      

CERTIFICATE AUTHORITY   EXPIRES                  RESIDUAL TIME   EXTERNALLY MANAGED
ca                      Oct 22, 2032 07:40 UTC   99y             no
etcd-ca                 Oct 22, 2032 07:40 UTC   99y             no
front-proxy-ca          Oct 22, 2032 07:40 UTC   99y             no   

# 查看kubeadm初始化控制平面配置信息
$ kubeadm config print init-defaults
apiVersion: kubeadm.k8s.io/v1beta3
bootstrapTokens:
- groups:
  - system:bootstrappers:kubeadm:default-node-token
  token: abcdef.0123456789abcdef
  ttl: 24h0m0s
  usages:
  - signing
  - authentication
kind: InitConfiguration
localAPIEndpoint:
  advertiseAddress: 1.2.3.4
  bindPort: 6443
nodeRegistration:
  criSocket: /var/run/dockershim.sock
  imagePullPolicy: IfNotPresent
  name: node
  taints: null
---
apiServer:
  timeoutForControlPlane: 4m0s
apiVersion: kubeadm.k8s.io/v1beta3
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
controllerManager: {}
dns: {}
etcd:
  local:
    dataDir: /var/lib/etcd
imageRepository: k8s.gcr.io
kind: ClusterConfiguration
kubernetesVersion: 1.22.0
networking:
  dnsDomain: cluster.local
  serviceSubnet: 10.96.0.0/12
scheduler: {}

# 查看kube-system空间Pod运行状态
$ kubectl get pod --namespace=kube-system
NAME                                       READY   STATUS             RESTARTS      AGE
calico-kube-controllers-86c9c65c67-j7pv4   1/1     Running            0             47m
calico-node-8mzpk                          1/1     Running            0             47m
calico-node-tkzs2                          1/1     Running            0             47m
calico-node-xbwvp                          1/1     Running            0             47m
coredns-7f6cbbb7b8-96hp9                   1/1     Running            0             82m
coredns-7f6cbbb7b8-kfmnn                   1/1     Running            0             82m
etcd-k8s-master01                          1/1     Running            0             82m
etcd-k8s-master02                          1/1     Running            0             72m
etcd-k8s-master03                          1/1     Running            0             70m
haproxy-k8s-master01                       1/1     Running            0             36m
haproxy-k8s-master02                       1/1     Running            0             67m
haproxy-k8s-master03                       1/1     Running            0             66m
keepalived-k8s-master01                    1/1     Running            0             82m
keepalived-k8s-master02                    1/1     Running            0             67m
keepalived-k8s-master03                    1/1     Running            0             66m
kube-apiserver-k8s-master01                1/1     Running            0             82m
kube-apiserver-k8s-master02                1/1     Running            0             72m
kube-apiserver-k8s-master03                1/1     Running            0             70m
kube-controller-manager-k8s-master01       1/1     Running            0             23m
kube-controller-manager-k8s-master02       1/1     Running            0             23m
kube-controller-manager-k8s-master03       1/1     Running            0             23m
kube-proxy-cvdlr                           1/1     Running            0             70m
kube-proxy-gnl7t                           1/1     Running            0             72m
kube-proxy-xnrt7                           1/1     Running            0             82m
kube-scheduler-k8s-master01                1/1     Running            0             23m
kube-scheduler-k8s-master02                1/1     Running            0             23m
kube-scheduler-k8s-master03                1/1     Running            0             23m
metrics-server-5786d84b7c-5v4rv            1/1     Running            0             8m38s