1 k8s 高可用集群架构

下面是 kubernetes 官网的集群架构图
在这里插入图片描述

2 部署 k8s 集群

2.1 集群规划

hostnameipcomponentscluster roleflannel versionkubectl versionkubeadm versionkubectl versionkeepalivedOSdocker versiondocker-root-datacgroup driver
www.datang001.com10.176.10.20kube-apiservermaster01v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
www.datang002.com10.176.10.21kube-apiservermaster02v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
www.datang003.com10.176.10.22kube-apiservermaster03v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
www.datang004.com10.176.10.23kubelet/kube-proxynode01v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
www.datang005.com10.176.10.24kubelet/kube-proxynode02v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
www.datang006.com10.176.10.25kubelet/kube-proxynode03v1.18.20v1.18.20v1.18.20Centos7.920.10.2/var/lib/dockersystemd
apiserver-lb.com10.176.10.250VIP

2.2 高可用架构

The kubeadm method is used to build a high-availability k8s cluster. The high availability of the k8s cluster is actually the high availability of the core components of k8s. This deployment adopts the active-standby mode. The architecture is as follows:
在这里插入图片描述

Description of the high-availability architecture in active-standby mode:

core componentshigh availablity modehigh availablity implement method
apiservermaster-backupkeepalived-+haproxy
controller-managermaster-backupleader election
schedulermaster-backupleader election
etcdclusterkubeadm
  • apiserver: High availability through keepalived, triggering keepalived vip transfer when a node fails
  • controller-manager: k8s generates a leader by election (controlled by --leader-elect, the default is true), and only one controller-manager component runs in the cluster at the same time;
  • scheduler: k8s generates a leader by election (controlled by --leader-elect, the default is true), and only one scheduler component runs in the cluster at the same time;
  • etcd:The cluster is automatically created by running kubeadm to achieve high availability. The number of deployed nodes is odd, and the 3-node mode tolerates at most one machine downtime.

2.3 开始部署

2.3.1 在每台机器 /etc/hosts 文件添加 IP 和主机名的映射

Execute the commad in all the control plan and work node hosts’s /etc/hosts file

cat >> /etc/hosts <<EOF
 
10.176.10.20 www.datang001.com
10.176.10.21 www.datang001.com
10.176.10.22 www.datang001.com
 
10.176.10.23 www.datang001.com
10.176.10.24 www.datang001.com
10.176.10.25 www.datang001.com
 
10.176.10.250 apiserver-lb.com
EOF
2.3.2 禁用 swap 分区、关闭 firewalld、禁用SeLinux

临时禁用 swap 分区，机器重启后失效，所有机器都执行

[root@www.datang001.com ~]# swapoff -a
[root@www.datang001.com ~]# free -m
              total        used        free      shared  buff/cache   available
Mem:         128770        4073      120315         330        4381      123763
Swap:             0           0           0

永久禁用 swap 分区，所有机器都执行

[root@www.datang001.com ~]# cat /etc/fstab
 
#
# /etc/fstab
# Created by anaconda on Thu Oct  5 04:55:59 2017
#
# Accessible filesystems, by reference, are maintained under '/dev/disk'
# See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info
#
...
...
#/dev/mapper/rootvg-swap swap                    swap    defaults        0 0
...

禁用 seLinux
首先查看 selinux 状态

[root@www.datang001.com ~]# sestatus -v
SELinux status:                 disabled

临时禁用

setenforce 0

永久禁用

sed -i 's/SELINUX=enforcing/SELINUX=disabled/g' /etc/selinux/config
2.3.3 升级内核和启用某些内核模块

为了集群的稳定性和防止后面业务容器把节点内存耗尽问题，生产环境必须升级linux服务器内核到4.19之上。由于centos7默认的内核版本是3.10.x，实际运行中，可能会出现内存泄露的问题，根因是 cgroup 的 keme account 特性有 内存泄露问题，具体分析请移步这里低内核造成k8s内存泄露，所以部署k8s集群之前，一定先对所有机器的内核版本进行升级。[已在使用的生产环境不要立刻做升级操作，因为已经有业务在kuberentes上运行了，升级的内核会导致业务容器飘逸，严重情况可能造成业务容器不能正常运行。]
内核升级过程:

略

转发 IPv4 并让 iptables 看到桥接流量

通过运行 lsmod | grep br_netfilter 来验证 br_netfilter 模块是否已加载。

若要显式加载此模块，请运行 sudo modprobe br_netfilter。

为了让 Linux 节点的 iptables 能够正确查看桥接流量，请确认 sysctl 配置中的 net.bridge.bridge-nf-call-iptables 设置为 1。例如：

cat <<EOF | sudo tee /etc/modules-load.d/k8s.conf
overlay
br_netfilter
EOF

sudo modprobe overlay
sudo modprobe br_netfilter

# 设置所需的 sysctl 参数，参数在重新启动后保持不变
cat <<EOF | sudo tee /etc/sysctl.d/k8s.conf
net.bridge.bridge-nf-call-iptables  = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.ipv4.ip_forward                 = 1
EOF

# 应用 sysctl 参数而不重新启动
sudo sysctl --system
2.3.4 设置 kubernetes repo 源

国内的一般设置为阿里云的源，如果服务器可以科学上网，那么直接使用默认谷歌源
不能科学上网，使用阿里镜像源

cat <<EOF > /etc/yum.repos.d/kubernetes.repo
[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

能科学上网，使用谷歌镜像源

cat <<EOF > /etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=https://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64
enabled=1
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
  • [] 中括号中的是repository id，唯一，用来标识不同仓库
  • name 仓库名称，自定义
  • baseurl 仓库地址
  • enable 是否启用该仓库，默认为1表示启用
  • gpgcheck 是否验证从该仓库获得程序包的合法性，1为验证
  • repo_gpgcheck 是否验证元数据的合法性 元数据就是程序包列表，1为验证
  • gpgkey=URL 数字签名的公钥文件所在位置，如果gpgcheck值为1，此处就需要指定gpgkey文件的位置，如果gpgcheck值为0就不需要此项了
2.3.4 在线安装 docker 和 kubeadm、kubelet、和 kubectl

首先升级 yum cache

yum clean all 
yum update
yum -y makecahe

先安装 docker，关于docker 的安装我们需要注意，docker 存储目录要爆炸尽量是大磁盘，因为后期我们需要拉取很多镜像文件呢，另外 Cgroup Driver 应该设置为 systemd， Storage Driver 应该设置为 overlay2。
/etc/docker/daemon.json

{
	"exec-opts": ["native.cgroupdriver=systemd"]，
	"data-root": "/data01/docker"
}

安装指定版本的docker

yum list docker-ce --showduplicates | sort -r
yum install -y docker-ce-19.03.9-3.el7
systemctl start docker && systemctl enable docker

安装指定版本的 kubeadm、kubelet、kubectl

yum -y install kubeadm==1.18.20 kubelet==1.18.20 kubectl==1.18.20
2.3.5 离线安装 docker 和 kubeadm、kubelet、和 kubectl，以及提前下载镜像

有时候我们机器处于内网中，不能正常连接互联网，这时我们就需要准备好 rpm 包来进行安装。这些 rpm 包是有相互依赖关系的，所以在安装的时候有先后顺序。kubectl 依赖于 crit-tools，kubernetes-cni 和 kubelet 之间相互依赖，kubeadm 依赖于 kubectl、kubelet和 crit-tools。

[root@www.datang001.com rpm]# pwd
/home/shutang/k8s/rpm
[root@www.datang001.com rpm]# ls
cri-tools-1.19.0-0.x86_64.rpm  kubeadm-1.18.20-0.x86_64.rpm  kubectl-1.18.20-0.x86_64.rpm  kubelet-1.18.20-0.x86_64.rpm  kubernetes-cni-0.8.7-0.x86_64.rpm
[root@www.datang001.com rpm]# yum -y install ./cri-tools-1.19.0-0.x86_64.rpm
[root@www.datang001.com rpm]# yum -y install ./kubectl-1.18.20-0.x86_64.rpm
[root@www.datang001.com rpm]# yum -y install ./kubernetes-cni-0.8.7-0.x86_64.rpm ./kubelet-1.18.20-0.x86_64.rpm
[root@www.datang001.com rpm]# yum -y install ./kubeadm-1.18.20-0.x86_64.rpm

[root@www.datang001.com rpm]# whereis kubeadm
kubeadm: /usr/bin/kubeadm
[root@www.datang001.com rpm]# whereis kubelet
kubelet: /usr/bin/kubelet
[root@www.datang001.com rpm]# whereis kubectl
kubectl: /usr/bin/kubectl

查看我们需要提前下载的镜像

[root@phx11-gliws-u23 ~]# kubeadm config images list --kubernetes-version v1.18.20
W1112 20:10:37.628119   20654 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io]
k8s.gcr.io/kube-apiserver:v1.18.20
k8s.gcr.io/kube-controller-manager:v1.18.20
k8s.gcr.io/kube-scheduler:v1.18.20
k8s.gcr.io/kube-proxy:v1.18.20
k8s.gcr.io/pause:3.2
k8s.gcr.io/etcd:3.4.3-0
k8s.gcr.io/coredns:1.6.7

# 注意：master节点需要把上面的镜像都下载，node节点只需要下载 k8s.gcr.io/kube-proxy:v1.18.20 k8s.gcr.io/pause:3.2 k8s.gcr.io/coredns:1.6.7

如果我们不能访问 k8s.gcr.io，我们需要利用阿里云提供的镜像仓库里的镜像。只不过有时候阿里云镜像仓库保存的版本与 k8s.gcr.io 不同步而已，如果需要安装比较新的版本，阿里云镜像仓库不存在的话，就可以用 daocloud 提供的镜像仓库或 清华镜像仓库，或者试试国内别的镜像仓库。

docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/kube-apiserver:v1.18.20
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/kube-controller-manager:v1.18.20
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/kube-scheduler:v1.18.20
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/kube-proxy:v1.18.20
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.2
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/etcd:3.4.3-0
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/coredns:1.6.7
2.3.6 设置 kubelet

默认配置的 pause 镜像使用 k8s.gcr.io 仓库，国内可能无法访问，所以这里配置 kubelet 使用阿里云的 pause 镜像地址。

DOCKER_CGROUPS=$(docker info |grep 'Cgroup' |cut -d '' -f4)
cat > /etc/sysconfig/kubelet <EOF
KUBELET_EXTRA_ARGS="--cgroup-dirver=$DOCKER_CGROUPS --pod-infrs-contaienr-image=registry.cn-hangzhou.aliyuncs.com/google_containers/puase-amd64:3.2"
EOF

设置 kubelet 开机自启动

systemctl daemon-reload
systemctl enable --now kubelet
2.3.7 集群初始化

https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/high-availability/
https://kubernetes.io/zh-cn/docs/reference/setup-tools/kubeadm/kubeadm-config/

在 master01 节点的 kubeadm-config.yaml 配置文件如下：

apiVersion: kubeadm.k8s.io/v1beta2
apiServer:
  certSANs:
  - apiserver-lb.com
  - www.datang001.com
  - www.datang002.com
  - www.datang003.com
  - www.datang004.com
  - www.datang005.com
  - www.datang006.com
  - 10.172.10.20
  - 10.172.10.21
  - 10.172.10.22
  - 10.172.10.23
  - 10.172.10.24
  - 10.172.10.25
  - 10.172.10.250
  extraArgs:
    authorization-mode: Node,RBAC
  timeoutForControlPlane: 4m0s
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
controlPlaneEndpoint: apiserver-lb:16443  # 修改成负载均衡的地址
controllerManager: {}
dns:
  type: CoreDNS
etcd:
  local:
    dataDir: /var/lib/etcd
imageRepository: k8s.gcr.io
#imageRepository: registory.cn-hangzhou.aliyuncs.com/google_container
#imageRepository: daocloud.io/daocloud
kind: ClusterConfiguration
kubernetesVersion: v1.18.20
networking:
  dnsDomain: cluster.local
  podSubnet: 172.26.0.0/16
  serviceSubnet: 10.96.0.0/12
scheduler: {}

后面我们升级kubenetes 版本的时候，再次初始化，可能 kubeadm-config 文件里的一些 API 有所变更，需要重新根据 old 的配置文件生产新的 kubeadm-config 文件

kubeadm config migrate --old-config kubeadm-config.yaml --new-config new.yaml

所有节点提前下载镜像，可以节省初始化时间

kubeadm config image pull --config kubeadm-config.yaml

master01 节点初始化，初始化以后会在 /etc/kubernetes 目录下生成对应的证书和配置文件，--upload-certs 参数是当有节点加入集群时，自动同步 master01 上生成的证书到该节点上。:

[root@www.datang001.com k8s]# sudo kubeadm init --config kubeadm.yaml --upload-certs
W0514 23:06:11.417640   20494 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io]
[init] Using Kubernetes version: v1.18.20
[preflight] Running pre-flight checks
    [WARNING SystemVerification]: this Docker version is not on the list of validated versions: 20.10.2. Latest validated version: 19.03
[preflight] Pulling images required for setting up a Kubernetes cluster
[preflight] This might take a minute or two, depending on the speed of your internet connection
[preflight] You can also perform this action in beforehand using 'kubeadm config images pull'
[kubelet-start] Writing kubelet environment file with flags to file "/var/lib/kubelet/kubeadm-flags.env"
[kubelet-start] Writing kubelet configuration to file "/var/lib/kubelet/config.yaml"
[kubelet-start] Starting the kubelet
[certs] Using certificateDir folder "/etc/kubernetes/pki"
[certs] Generating "ca" certificate and key
[certs] Generating "apiserver" certificate and key
[certs] apiserver serving cert is signed for DNS names [www.datang001.com kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local apiserver-lb.com] and IPs [10.96.0.1 10.222.175.201]
[certs] Generating "apiserver-kubelet-client" certificate and key
[certs] Generating "front-proxy-ca" certificate and key
[certs] Generating "front-proxy-client" certificate and key
[certs] Generating "etcd/ca" certificate and key
[certs] Generating "etcd/server" certificate and key
[certs] etcd/server serving cert is signed for DNS names [phx11-gliws-u23 localhost] and IPs [10.172.10.20 127.0.0.1 ::1]
[certs] Generating "etcd/peer" certificate and key
[certs] etcd/peer serving cert is signed for DNS names [www.datang001.com localhost] and IPs [10.172.10.20 127.0.0.1 ::1]
[certs] Generating "etcd/healthcheck-client" certificate and key
[certs] Generating "apiserver-etcd-client" certificate and key
[certs] Generating "sa" key and public key
[kubeconfig] Using kubeconfig folder "/etc/kubernetes"
[kubeconfig] Writing "admin.conf" kubeconfig file
[kubeconfig] Writing "kubelet.conf" kubeconfig file
[kubeconfig] Writing "controller-manager.conf" kubeconfig file
[kubeconfig] Writing "scheduler.conf" kubeconfig file
[control-plane] Using manifest folder "/etc/kubernetes/manifests"
[control-plane] Creating static Pod manifest for "kube-apiserver"
[control-plane] Creating static Pod manifest for "kube-controller-manager"
W0514 23:06:16.003004   20494 manifests.go:225] the default kube-apiserver authorization-mode is "Node,RBAC"; using "Node,RBAC"
[control-plane] Creating static Pod manifest for "kube-scheduler"
W0514 23:06:16.004606   20494 manifests.go:225] the default kube-apiserver authorization-mode is "Node,RBAC"; using "Node,RBAC"
[etcd] Creating static Pod manifest for local etcd in "/etc/kubernetes/manifests"
[wait-control-plane] Waiting for the kubelet to boot up the control plane as static Pods from directory "/etc/kubernetes/manifests". This can take up to 4m0s
[apiclient] All control plane components are healthy after 20.502817 seconds
[upload-config] Storing the configuration used in ConfigMap "kubeadm-config" in the "kube-system" Namespace
[kubelet] Creating a ConfigMap "kubelet-config-1.18" in namespace kube-system with the configuration for the kubelets in the cluster
[upload-certs] Storing the certificates in Secret "kubeadm-certs" in the "kube-system" Namespace
[upload-certs] Using certificate key:
**************************************************
[mark-control-plane] Marking the node www.datang001.com as control-plane by adding the label "node-role.kubernetes.io/master=''"
[mark-control-plane] Marking the node www.datang001.com as control-plane by adding the taints [node-role.kubernetes.io/master:NoSchedule]
[bootstrap-token] Using token: ixhv5g.n37m33eybijtb13q
[bootstrap-token] Configuring bootstrap tokens, cluster-info ConfigMap, RBAC Roles
[bootstrap-token] configured RBAC rules to allow Node Bootstrap tokens to get nodes
[bootstrap-token] configured RBAC rules to allow Node Bootstrap tokens to post CSRs in order for nodes to get long term certificate credentials
[bootstrap-token] configured RBAC rules to allow the csrapprover controller automatically approve CSRs from a Node Bootstrap Token
[bootstrap-token] configured RBAC rules to allow certificate rotation for all node client certificates in the cluster
[bootstrap-token] Creating the "cluster-info" ConfigMap in the "kube-public" namespace
[kubelet-finalize] Updating "/etc/kubernetes/kubelet.conf" to point to a rotatable kubelet client certificate and key
[addons] Applied essential addon: CoreDNS
[addons] Applied essential addon: kube-proxy
 
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
 
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 apiserver-lb.com:6443 --token ixhv5g.n37m33eybijtb13q \
    --discovery-token-ca-cert-hash sha256:fc9a9ff3fc5ae118a5a9616cb742a26deacc235ec79beb85018b52280d887d5e \
    --control-plane --certificate-key *************************************************
 
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 apiserver-lb.com:6443 --token ixhv5g.n37m33eybijtb13q \
    --discovery-token-ca-cert-hash sha256:fc9a9ff3fc5ae118a5a9616cb742a26deacc235ec79beb85018b52280d887d5e
[root@www.datang001.com k8s]# mkdir -p $HOME/.kube
[root@www.datang001.com k8s]# sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
[root@www.datang001.com k8s]# sudo chown $(id -u):$(id -g) $HOME/.kube/config

或者不用指定配置文件初始化：

kubeadm init --control-plane-endpoint "LOAD_BALANCER_DNS:LOAD_BALANCER_PORT" --image-repository daocloud.io/daocloud --upload-certs

如果初始化失败，重置后再次初始化，命令如下：

kubeadm reset

Token 过期处理：https://kubernetes.io/zh-cn/docs/reference/setup-tools/kubeadm/kubeadm-token/#cmd-token-create

2.3.8 把其他 master 节点和 node 节点加入集群

master 节点加入集群

kubeadm join apiserver-lb.com:6443 --token ixhv5g.n37m33eybijtb13q \
    --discovery-token-ca-cert-hash sha256:fc9a9ff3fc5ae118a5a9616cb742a26deacc235ec79beb85018b52280d887d5e \
    --control-plane --certificate-key *************************************************

node 节点加入集群

kubeadm join apiserver-lb.com:6443 --token ixhv5g.n37m33eybijtb13q \
    --discovery-token-ca-cert-hash sha256:fc9a9ff3fc5ae118a5a9616cb742a26deacc235ec79beb85018b52280d887d5e
2.3.9 在 master01 节点上安装 keepalived、haproxy 软件
yum -y install keepalived haproxy

master01 机器上的 keepalived.conf 配置文件 和 check_apiserver.sh 脚本文件。

# keepalived.conf 内容
! Configuration File for keepalived
global_defs {
   router_id www.datang001.com
}

# 定义脚本
vrrp_script check_apiserver {
    script "/etc/keepalived/check_apiserver.sh" 
    interval 2                                  
    weight -5                                  
    fall 3                                   
    rise 2                               
}

vrrp_instance VI_1 {
    state MASTER 
    interface eth0
    virtual_router_id 50
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.176.10.250
    }

    # 调用脚本
    track_script {
        check_apiserver
    }
}


# 监测脚本 check_apiserver
#!/bin/bash

function check_apiserver(){
  for ((i=0;i<5;i++))
  do
    apiserver_job_id=${pgrep kube-apiserver}
    if [[ ! -z ${apiserver_job_id} ]];then
      return
    else
      sleep 2
    fi
  done
  apiserver_job_id=0
}

# 1->running    0->stopped
check_apiserver
if [[ $apiserver_job_id -eq 0 ]];then
  /usr/bin/systemctl stop keepalived
  exit 1
else
  exit 0
fi

启动 keepalived

systemctl enable --now keepalived.service

master01 上 haproxy 的配置文件 haproxy.cfg。

global
    log /dev/log  local0 warning
    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user        haproxy
    group       haproxy
    daemon

   stats socket /var/lib/haproxy/stats

defaults
  mode http
  log global
  option  httplog
  option  dontlognull
        timeout connect 5000
        timeout client 50000
        timeout server 50000

listen status_page
    bind 0.0.0.0:1080
    stats enable
    stats uri /haproxy-status
    stats auth    admin:nihaoma
    stats realm "Welcome to the haproxy load balancer status page"
    stats hide-version
    stats admin if TRUE
    stats refresh 5s

frontend kube-apiserver
  bind *:16443
  mode tcp
  option tcplog
  default_backend kube-apiserver

backend kube-apiserver
    mode tcp
    option tcplog
    option tcp-check
    balance roundrobin
    default-server inter 10s downinter 5s rise 2 fall 2 slowstart 60s maxconn 250 maxqueue 256 weight 100
    server www.datang001.com           10.172.10.20:6443  check # Replace the IP address with your own.
    server www.datang002.com           10.172.10.21:6443 check 
    server www.datang003.com           10.172.110.22:6443  check

启动 haproxy

systemctl enable --now haproxy

master02 机器上的 keepalived.conf 配置文件 和 check_apiserver.sh 脚本文件。

! Configuration File for keepalived
global_defs {
   router_id www.datang002.com
}

# 定义脚本
vrrp_script check_apiserver {
    script "/etc/keepalived/check_apiserver.sh"
    interval 2
    weight -5
    fall 3
    rise 2
}

vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 50
    priority 99
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.172.10.250
    }

    # 调用脚本
    #track_script {
    #    check_apiserver
    #}
}

master02 上 haproxy 的配置文件 haproxy.cfg。

global
    log /dev/log  local0 warning
    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user        haproxy
    group       haproxy
    daemon

   stats socket /var/lib/haproxy/stats

defaults
  mode http
  log global
  option  httplog
  option  dontlognull
        timeout connect 5000
        timeout client 50000
        timeout server 50000

listen status_page
    bind 0.0.0.0:1080
    stats enable
    stats uri /haproxy-status
    stats auth    admin:nihaoma
    stats realm "Welcome to the haproxy load balancer status page"
    stats hide-version
    stats admin if TRUE
    stats refresh 5s

frontend kube-apiserver
  bind *:16443
  mode tcp
  option tcplog
  default_backend kube-apiserver

backend kube-apiserver
    mode tcp
    option tcplog
    option tcp-check
    balance roundrobin
    default-server inter 10s downinter 5s rise 2 fall 2 slowstart 60s maxconn 250 maxqueue 256 weight 100
    server www.datang001.com           10.172.10.20:6443  check # Replace the IP address with your own.
    server www.datang002.com           10.172.10.21:6443 check 
    server www.datang003.com           10.172.110.22:6443  check

master03 机器上的 keepalived.conf 配置文件 和 check_apiserver.sh 脚本文件。
master03 上 haproxy 的配置文件 haproxy.cfg。

2.3.10 安装 flannel 网络
[root@www.datang001.com network]$ wget https://raw.githubusercontent.com/coreos/flannel/2140ac876ef134e0ed5af15c65e414cf26827915/Documentation/kube-flannel.yml
[root@www.datang001.com network]$ kubectl apply -f kube-flannel.yml
2.3.11 查看集群状态
[shutang@www.datang001.com network]# kubectl get nodes -o wide
NAME              STATUS   ROLES    AGE    VERSION    INTERNAL-IP      EXTERNAL-IP   OS-IMAGE                KERNEL-VERSION                CONTAINER-RUNTIME
www.datang001.com   Ready    <none>   160m   v1.18.20   10.172.10.20    <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.2
www.datang002.com   Ready    <none>   160m   v1.18.20   10.172.10.21   <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.2
www.datang003.com   Ready    <none>   161m   v1.18.20   10.172.10.22   <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.2
www.datang004.com   Ready    master   162m   v1.18.20   10.172.10.23   <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.2
www.datang005.com   Ready    master   163m   v1.18.20   10.172.10.24   <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.6
www.datang006.com   Ready    master   166m   v1.18.20   10.172.10.25   <none>        CentOS Linux 7 (Core)   3.10.0-1160.62.1.el7.x86_64   docker://20.10.2



# execute kubectl get cs It may be that the first two components are unhealthy in reality, and there are ways to deal with it later in the article
[root@www.datang001.com ~]# kubectl get cs
NAME                 STATUS    MESSAGE             ERROR
controller-manager   Healthy   ok
scheduler            Healthy   ok
etcd-0               Healthy   {"health":"true"}
2.3.12 修改 k8s 默认的 NodePort 端口范围

In a kubernetes cluster, the default range of NodePort is 30000-32767. In some cases, due to company network policy restrictions, you may modify the port range of NodePort.

Modify kube-apiserver.yaml

When using kubeadm to install a k8s cluster, there will be a file /etc/kubernetes/mainfests/kube-apiserver.yaml on your control plane nodes, modify this file and add --service-node-port-range=1 to it -65535 (please use your own desired port range) as follows:

apiVersion: v1
kind: Pod
metadata:
  annotations:
    kubeadm.kubernetes.io/kube-apiserver.advertise-address.endpoint: 10.172.10.20:6443
  creationTimestamp: null
  labels:
    component: kube-apiserver
    tier: control-plane
  name: kube-apiserver
  namespace: kube-system
spec:
  containers:
  - command:
    - kube-apiserver
    - --advertise-address=10.172.10.20
    - --allow-privileged=true
    - --authorization-mode=Node,RBAC
    - --client-ca-file=/etc/kubernetes/pki/ca.crt
    - --enable-admission-plugins=NodeRestriction
    - --enable-bootstrap-token-auth=true
    - --etcd-cafile=/etc/kubernetes/pki/etcd/ca.crt
    - --etcd-certfile=/etc/kubernetes/pki/apiserver-etcd-client.crt
    - --etcd-keyfile=/etc/kubernetes/pki/apiserver-etcd-client.key
    - --etcd-servers=https://127.0.0.1:2379
    - --insecure-port=0
    - --kubelet-client-certificate=/etc/kubernetes/pki/apiserver-kubelet-client.crt
    - --kubelet-client-key=/etc/kubernetes/pki/apiserver-kubelet-client.key
    - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
    - --proxy-client-cert-file=/etc/kubernetes/pki/front-proxy-client.crt
    - --proxy-client-key-file=/etc/kubernetes/pki/front-proxy-client.key
    - --requestheader-allowed-names=front-proxy-client
    - --requestheader-client-ca-file=/etc/kubernetes/pki/front-proxy-ca.crt
    - --requestheader-extra-headers-prefix=X-Remote-Extra-
    - --requestheader-group-headers=X-Remote-Group
    - --requestheader-username-headers=X-Remote-User
    - --secure-port=6443
    - --service-account-key-file=/etc/kubernetes/pki/sa.pub
    - --service-cluster-ip-range=10.96.0.0/12
    - --service-node-port-range=1-65535                        # 新增该行
    - --tls-cert-file=/etc/kubernetes/pki/apiserver.crt
    - --tls-private-key-file=/etc/kubernetes/pki/apiserver.key
    image: k8s.gcr.io/kube-apiserver:v1.18.20
    imagePullPolicy: IfNotPresent
    ......

重启 apiserver

# get apiserver pod name
export apiserver_pods=$(kubectl get pods --selector=component=kube-apiserver -n kube-system --output=jsonpath={.items..metadata.name})
# delete apiserver pod
kubectl delete pod $apiserver_pods -n kube-system

监测 apiserver 是否正常

kubectl describe pod $apiserver_pods -n kube-system
 
# Check whether there is the line we added above in the parameters of the startup command, and if so, verify that it is correct
2.3.13 修改集群证书时间
2.3.13.1 First of all, you need to determine the kubernetes version used in the installation of the k8s cluster and the go version when developing this version
[root@www.datang001.com ~]# kubeadm version
kubeadm version: &version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.20", GitCommit:"1f3e19b7beb1cc0110255668c4238ed63dadb7ad", GitTreeState:"clean", BuildDate:"2021-06-16T12:56:41Z", GoVersion:"go1.13.15", Compiler:"gc", Platform:"linux/amd64"}
2.3.13.2 Below we download the kubernetes1.18.20.tar.gz package and install the go1.13.15 tool
[root@www.datang001.com update-cert]# wget https://github.com/kubernetes/kubernetes/archive/refs/tags/v1.18.20.tar.gz && wget https://golang.google.cn/dl/go1.13.15.linux-amd64.tar.gz
[root@www.datang001.com update-cert]# tar -zxf v1.18.20.tar.gz && tar -zxf go1.13.15.linux-amd64.tar.gz -C /usr/local/
 
[root@www.datang001.com update-cert]# cat > /etc/profile.d/go.sh <<EOF
export PATH=$PATH:/usr/local/go/bin
EOF
[root@www.datang001.com update-cert]# source /etc/profile.d/go.sh
[root@www.datang001.com update-cert]# go version
go version go1.13.15 linux/amd64
[root@www.datang001.com update-cert]#
2.3.13.3 Using k8s installed by kubeadm, all certificates are placed in the directory /etc/kubernetes/pki. We can check the time of each certificate and find that the certificate of the ca class is valid for 10 years, in addition to other components Certificates are valid for one year by default.
[root@www.datang001.com ~]# kubeadm alpha certs check-expiration
[check-expiration] Reading configuration from the cluster...
[check-expiration] FYI: You can look at this config file with 'kubectl -n kube-system get cm kubeadm-config -oyaml'
 
CERTIFICATE                EXPIRES                  RESIDUAL TIME   CERTIFICATE AUTHORITY   EXTERNALLY MANAGED
admin.conf                 May 15, 2023 06:06 UTC   364d                                    no
apiserver                  May 15, 2023 06:06 UTC   364d            ca                      no
apiserver-etcd-client      May 15, 2023 06:06 UTC   364d            etcd-ca                 no
apiserver-kubelet-client   May 15, 2023 06:06 UTC   364d            ca                      no
controller-manager.conf    May 15, 2023 06:06 UTC   364d                                    no
etcd-healthcheck-client    May 15, 2023 06:06 UTC   364d            etcd-ca                 no
etcd-peer                  May 15, 2023 06:06 UTC   364d            etcd-ca                 no
etcd-server                May 15, 2023 06:06 UTC   364d            etcd-ca                 no
front-proxy-client         May 15, 2023 06:06 UTC   364d            front-proxy-ca          no
scheduler.conf             May 15, 2023 06:06 UTC   364d                                    no
 
CERTIFICATE AUTHORITY   EXPIRES                  RESIDUAL TIME   EXTERNALLY MANAGED
ca                      May 12, 2032 06:06 UTC   9y              no
etcd-ca                 May 12, 2032 06:06 UTC   9y              no
front-proxy-ca          May 12, 2032 06:06 UTC   9y              no
2.3.13.4 Modify the corresponding constant source code related to the certificate time

www.datang001.com modified certs validate time

[root@www.datang001.com update-cert]# pwd
/home/shutang/k8s/update-cert
[root@www.datang001.com update-cert]# cd kubernetes-1.18.20
[root@www.datang001.com kubernetes-1.18.20]# cd cmd/kubeadm/app/constants/
[root@www.datang001.com constants]# cat constants |grep 10
cat: constants: No such file or directory
[root@www.datang001.com constants]# cat constants.go |grep 10
    CertificateValidity = time.Hour * 24 * 365 * 10
 
[root@www.datang001.com kubernetes-1.18.20]# make WHAT=cmd/kubeadm
+++ [0515 08:40:42] Building go targets for linux/amd64:
    ./vendor/k8s.io/code-generator/cmd/deepcopy-gen
+++ [0515 08:40:52] Building go targets for linux/amd64:
    ./vendor/k8s.io/code-generator/cmd/defaulter-gen
+++ [0515 08:40:59] Building go targets for linux/amd64:
    ./vendor/k8s.io/code-generator/cmd/conversion-gen
+++ [0515 08:41:11] Building go targets for linux/amd64:
    ./vendor/k8s.io/kube-openapi/cmd/openapi-gen
+++ [0515 08:41:22] Building go targets for linux/amd64:
    ./vendor/github.com/go-bindata/go-bindata/go-bindata
warning: ignoring symlink /home/shutang/k8s/update-cert/kubernetes-1.18.20/_output/local/go/src/k8s.io/kubernetes
go: warning: "k8s.io/kubernetes/vendor/github.com/go-bindata/go-bindata/..." matched no packages
+++ [0515 08:41:24] Building go targets for linux/amd64:
    cmd/kubeadm
 
# backup the old kubeadm
[root@www.datang001.com kubernetes-1.18.20]# mv /usr/bin/kubeadm /usr/bin/kubeadm.old
[root@www.datang001.com kubernetes-1.18.20]# cp _output/bin/kubeadm /usr/bin/kubeadm
[root@www.datang001.com kubernetes-1.18.20]# cd /etc/kubernetes/pki/
[root@www.datang001.com pki]# ls -lah
total 60K
drwxr-xr-x 3 root root 4.0K May 14 23:06 .
drwxr-xr-x 4 root root  125 May 14 23:06 ..
-rw-r--r-- 1 root root 1.3K May 14 23:06 apiserver.crt
-rw-r--r-- 1 root root 1.1K May 14 23:06 apiserver-etcd-client.crt
-rw------- 1 root root 1.7K May 14 23:06 apiserver-etcd-client.key
-rw------- 1 root root 1.7K May 14 23:06 apiserver.key
-rw-r--r-- 1 root root 1.1K May 14 23:06 apiserver-kubelet-client.crt
-rw------- 1 root root 1.7K May 14 23:06 apiserver-kubelet-client.key
-rw-r--r-- 1 root root 1.1K May 14 23:06 ca.crt
-rw------- 1 root root 1.7K May 14 23:06 ca.key
drwxr-xr-x 2 root root  162 May 14 23:06 etcd
-rw-r--r-- 1 root root 1.1K May 14 23:06 front-proxy-ca.crt
-rw------- 1 root root 1.7K May 14 23:06 front-proxy-ca.key
-rw-r--r-- 1 root root 1.1K May 14 23:06 front-proxy-client.crt
-rw------- 1 root root 1.7K May 14 23:06 front-proxy-client.key
-rw------- 1 root root 1.7K May 14 23:06 sa.key
-rw------- 1 root root  451 May 14 23:06 sa.pub
[root@phx11-gliws-u23 pki]# kubeadm alpha certs renew all
[renew] Reading configuration from the cluster...
[renew] FYI: You can look at this config file with 'kubectl -n kube-system get cm kubeadm-config -oyaml'
 
certificate embedded in the kubeconfig file for the admin to use and for kubeadm itself renewed
certificate for serving the Kubernetes API renewed
certificate the apiserver uses to access etcd renewed
certificate for the API server to connect to kubelet renewed
certificate embedded in the kubeconfig file for the controller manager to use renewed
certificate for liveness probes to healthcheck etcd renewed
certificate for etcd nodes to communicate with each other renewed
certificate for serving etcd renewed
certificate for the front proxy client renewed
certificate embedded in the kubeconfig file for the scheduler manager to use renewed
[root@phx11-gliws-u23 pki]#
[root@phx11-gliws-u23 pki]# kubeadm alpha certs check-expiration
[check-expiration] Reading configuration from the cluster...
[check-expiration] FYI: You can look at this config file with 'kubectl -n kube-system get cm kubeadm-config -oyaml'
 
CERTIFICATE                EXPIRES                  RESIDUAL TIME   CERTIFICATE AUTHORITY   EXTERNALLY MANAGED
admin.conf                 May 12, 2032 15:46 UTC   9y                                      no
apiserver                  May 12, 2032 15:46 UTC   9y              ca                      no
apiserver-etcd-client      May 12, 2032 15:46 UTC   9y              etcd-ca                 no
apiserver-kubelet-client   May 12, 2032 15:46 UTC   9y              ca                      no
controller-manager.conf    May 12, 2032 15:46 UTC   9y                                      no
etcd-healthcheck-client    May 12, 2032 15:46 UTC   9y              etcd-ca                 no
etcd-peer                  May 12, 2032 15:46 UTC   9y              etcd-ca                 no
etcd-server                May 12, 2032 15:46 UTC   9y              etcd-ca                 no
front-proxy-client         May 12, 2032 15:46 UTC   9y              front-proxy-ca          no
scheduler.conf             May 12, 2032 15:46 UTC   9y                                      no
 
CERTIFICATE AUTHORITY   EXPIRES                  RESIDUAL TIME   EXTERNALLY MANAGED
ca                      May 12, 2032 06:06 UTC   9y              no
etcd-ca                 May 12, 2032 06:06 UTC   9y              no
front-proxy-ca          May 12, 2032 06:06 UTC   9y              no
2.3.14 修改 kube-proxy 的代理模式为 ipvs
2.3.14.1 查看 kubernetes 的kube-proxy 组件的代理模式
[root@www.datang001.com ~]# kubectl get pods -n kube-system |grep proxy
kube-proxy-2qc7r                          1/1     Running   13         182d
kube-proxy-6nfzm                          1/1     Running   13         182d
kube-proxy-frwcg                          1/1     Running   15         182d
kube-proxy-l6xg2                          1/1     Running   13         182d
kube-proxy-r96hz                          1/1     Running   13         182d
kube-proxy-sgwfh                          1/1     Running   13         182d

随便查看某个 pod 的日志

[shutang@phx11-gliws-u23 ~]$ kubectl logs -f kube-proxy-2qc7r -n kube-system
I1021 04:57:16.251139       1 node.go:136] Successfully retrieved node IP: 10.222.175.237
I1021 04:57:16.251166       1 server_others.go:259] Using iptables Proxier.
I1021 04:57:16.251599       1 server.go:583] Version: v1.18.20
I1021 04:57:16.252018       1 conntrack.go:52] Setting nf_conntrack_max to 131072
I1021 04:57:16.252241       1 config.go:133] Starting endpoints config controller
I1021 04:57:16.252270       1 shared_informer.go:223] Waiting for caches to sync for endpoints config
I1021 04:57:16.252300       1 config.go:315] Starting service config controller
I1021 04:57:16.252306       1 shared_informer.go:223] Waiting for caches to sync for service config
I1021 04:57:16.352464       1 shared_informer.go:230] Caches are synced for endpoints config
I1021 04:57:16.352528       1 shared_informer.go:230] Caches are synced for service config
E1103 07:57:22.505665       1 graceful_termination.go:89] Try delete

此时 kube-proxy 的默认代理模式 iptables。

2.3.14.2 设置 kube-proxy 代理模式为 ipvs

确保 ipvs 模式已经运行

[shutang@www.datang001.com ~]# lsmod |grep ip_vs
ip_vs_sh               12688  0
ip_vs_wrr              12697  0
ip_vs_rr               12600  167
ip_vs                 145458  173 ip_vs_rr,ip_vs_sh,ip_vs_wrr
nf_conntrack          139264  7 ip_vs,nf_nat,nf_nat_ipv4,xt_conntrack,nf_nat_masquerade_ipv4,nf_conntrack_netlink,nf_conntrack_ipv4
libcrc32c              12644  4 xfs,ip_vs,nf_nat,nf_conntrack

如果没有加载 ipvs 模块运行一下命令

cat > /etc/sysconfig/modules/ipvs.modules <<EOF
#!/bin/bash
modprobe -- ip_vs
modprobe -- ip_vs_rr
modprobe -- ip_vs_wrr
modprobe -- ip_vs_sh
modprobe -- nf_conntrack
EOF

执行下面命令

chmod 755 /etc/sysconfig/modules/ipvs.modules && bash /etc/sysconfig/modules/ipvs.modules && lsmod | grep -e ip_vs -e nf_conntrack_ipv4

修改 kube-proxy 的 configmap 文件

...
...
...
 kind: KubeProxyConfiguration
    metricsBindAddress: ""
    mode: "ipvs" #修改此处，原为空
    nodePortAddresses: null
    oomScoreAdj: null
    portRange: ""
...
...
...

重启 kube-proxy

kubectl rollout restart daemonset kube-proxy -n kube-system

2.4 安装集群web管理工具 kubesphere

3 排除故障总结

# kubernetes
Logo
K8S/Kubernetes

K8S/Kubernetes社区为您提供最前沿的新闻资讯和知识内容

更多推荐

【深度】阿里巴巴万级规模 K8s 集群全局高可用体系之美

作者 |  韩堂、柘远、沉醉来源 | 阿里巴巴云原生公众号​前言台湾作家林清玄在接受记者采访的时候，如此评价自己 30 多年写作生涯：“第一个十年我才华横溢，‘贼光闪现’，令周边黯然失色；第二个十年，我终于‘宝光现形’，不再去抢风头，反而与身边的美丽相得益彰；进入第三个十年，繁华落尽见真醇，我进入了‘醇光初现’的阶段，真正体味到了境界之美”。​长夜有穷，真水无香。领略过了 K8s“身在江

avatar K8S/Kubernetes

如何基于 K8s 构建下一代 DevOps 平台？

作者 | 孙健波（天元）导读：当前云原生 DevOps 体系现状如何？面临哪些挑战？如何通过 OAM 解决云原生 DevOps 场景下的诸多问题？云原生开发应用模型 OAM(Open Application Model) 社区核心成员孙健波将为大家一一解答，并分享如何基于 OAM 和 Kubernetes 打造无限能力的下一代 DevOps 平台。什么是 DevOps？为什么基于 Kub

avatar K8S/Kubernetes

k8s 火了！

2020，上云之年，产品云端化成为一种趋势。在一线城市，很多公司都已经构建了自己的私有云环境，比如阿里云、网易云、华为云等。而Kubernetes 作为基于容器编排领域的王者，具备扩展...

avatar K8S/Kubernetes