k8s集群角色分配
部署k8s集群的节点按照用途可以划分为如下2类角色:
- master:集群的master节点,集群的初始化节点,基础配置不低于2C4G
- slave:集群的worker节点,可以多台,基础配置不低于2C4G
安装指定版本k8s
可变参数
代码语言:javascript复制# 指定 Kubernetes 版本; 参考https://github.com/kubernetes/kubernetes/releases
kubernetes_version="1.27.3"
# 控制平面终结点地址,用于在高可用集群中指定负载均衡器的地址;选填
kubeadm init ...... --control-plane-endpoint=test-k8s-lb.opsbase.cn:6443
Master 执行脚本
- 自动判断国家切换镜像源
- 使用Containerd弃用Docker
- Centos7 & Ubuntu 22.4 & AWS Linux2
master节点:bash kubernets-install.sh master
worker节点:bash kubernets-install.sh node
代码语言:javascript复制#!/bin/bash
# 指定 Kubernetes 版本
kubernetes_version="1.27.4"
# 判断传递的参数,如果没有传递或传递的是错误参数,则默认安装master节点
node_type=${1:-"master"}
# 脚本用途说明
cat <<EOF
该脚本用于安装 Kubernetes 集群,并根据地区选择合适的镜像源。
请在运行脚本之前确认:
===================================================================
1. 安装集群Master节点: bash kubernets-install.sh master
2. 安装worker节点: kubernets-install.sh worker
3. 当前用户是 root 用户
4. 确保系统网络畅通,可以访问外部镜像源
5. 指定kubernetes安装版本
6. 默认使用flannel网络组件,可注释并改为install_network_plugin_calico
===================================================================
EOF
# 检查当前用户是否为 root 用户
check_root_user() {
if [[ $EUID -ne 0 ]]; then
echo "请使用 root 用户执行此脚本。"
exit 1
fi
}
# 判断是否为中国地区
is_china() {
# 使用简单的方法判断,您也可以根据实际需求添加更多判断条件
if [[ $(curl -sSL https://ipapi.co/country/) = "CN" ]]; then
return 0
else
return 1
fi
}
# 根据地区选择镜像源
select_country() {
if is_china; then
echo "检测在中国地区,将使用国内镜像源。"
docker_image_repository="registry.aliyuncs.com/google_containers"
yum_repository="https://mirrors.aliyun.com/kubernetes"
apt_repository="https://mirrors.aliyun.com/kubernetes/apt"
flannel="https://gitee.com/mirrors/flannel/raw/master/Documentation/kube-flannel.yml"
calico="https://docs.projectcalico.org/v3.20/manifests/calico.yaml --image-repository=registry.cn-hangzhou.aliyuncs.com/calico"
else
echo "检测不在中国地区,将使用官方镜像源。"
docker_image_repository="registry.k8s.io"
yum_repository="https://packages.cloud.google.com"
apt_repository="https://apt.kubernetes.io"
flannel="https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml"
calico="https://docs.projectcalico.org/v3.20/manifests/calico.yaml"
fi
}
# 检查是否已安装 Kubernetes
check_kubernetes_installed() {
if command -v kubeadm >/dev/null 2>&1; then
echo "已检测到已安装的 Kubernetes。"
read -p "是否卸载已存在的 Kubernetes?(y/n): " uninstall_choice
if [[ $uninstall_choice = "y" || $uninstall_choice = "Y" ]]; then
uninstall_kubernetes
else
echo "已取消安装。"
exit 0
fi
fi
}
# 卸载 Kubernetes
uninstall_kubernetes() {
echo "正在卸载 Kubernetes..."
case $os in
ubuntu)
uninstall_kubernetes_ubuntu
;;
centos)
uninstall_kubernetes_centos
;;
amazon_linux)
uninstall_kubernetes_centos
;;
*)
echo "不支持的操作系统。"
exit 1
;;
esac
echo "Kubernetes 已成功卸载。"
}
# 获取操作系统信息
get_os_info() {
if [ -f /etc/os-release ]; then
. /etc/os-release
if [[ $ID = "ubuntu" ]]; then
os="ubuntu"
elif [[ $ID = "centos" ]]; then
os="centos"
elif [[ $ID = "amzn" ]]; then
os="amazon_linux"
fi
elif [ -f /etc/redhat-release ]; then
if grep -q "CentOS Linux release 7" /etc/redhat-release; then
os="centos"
fi
fi
}
# 卸载 Kubernetes(Ubuntu)
uninstall_kubernetes_ubuntu() {
echo "正在卸载 Kubernetes..."
if command -v kubeadm &>/dev/null; then
kubeadm reset -f
else
echo "kubeadm 未找到,无法执行重置操作。请手动重置 Kubernetes。"
fi
if command -v kubectl &>/dev/null; then
kubectl delete -f $flannel
kubectl delete -f $calico
apt remove -y kubeadm kubelet kubectl containerd
rm -rf /etc/kubernetes /var/lib/etcd /var/lib/kubelet
else
echo "kubectl 未找到,无法执行删除操作。请手动删除相关资源。"
fi
}
# 卸载 Kubernetes(CentOS)
uninstall_kubernetes_centos() {
echo "正在卸载 Kubernetes..."
if command -v kubectl &>/dev/null; then
kubectl delete -f $flannel
kubectl delete -f $calico
yum --debuglevel=1 remove -y kubeadm kubelet kubectl containerd bash-completion
yum autoremove -y
rm -rf /etc/kubernetes /var/lib/etcd /var/lib/kubelet
else
echo "kubectl 未找到,无法执行删除操作。请手动删除相关资源。"
fi
}
# 关闭并禁用防火墙(Ubuntu、CentOS)
disable_firewall() {
echo "正在关闭并禁用防火墙..."
if [[ $os = "ubuntu" ]]; then
ufw disable
elif [[ $os = "centos" || $os = "amazon_linux" ]]; then
systemctl stop firewalld
systemctl disable firewalld
# 清空iptables策略
iptables -F
iptables -X
iptables -Z
iptables -F -t nat
iptables -X -t nat
iptables -Z -t nat
iptables -P INPUT ACCEPT
if [ -s /etc/selinux/config ]; then
setenforce 0
sed -i 's/^SELINUX=.*/SELINUX=disabled/g' /etc/selinux/config
fi
fi
}
# 关闭并禁用 Swap
disable_swap() {
echo "正在关闭并禁用 Swap..."
swapoff -a
sed -i '/swap/d' /etc/fstab
}
# 优化内核参数
optimize_kernel() {
echo "正在优化内核参数..."
sysctl_file="/etc/sysctl.d/kubernetes.conf"
echo "net.bridge.bridge-nf-call-ip6tables = 1" >$sysctl_file
echo "net.bridge.bridge-nf-call-iptables = 1" >>$sysctl_file
echo "net.ipv4.ip_forward=1" >>$sysctl_file
echo "vm.max_map_count=262144" >>$sysctl_file
sysctl -p $sysctl_file
}
# 禁用透明大页
disable_transparent_hugepage() {
echo "禁用透明大页..."
thp_file="/etc/systemd/system/disable-thp.service"
echo "[Unit]" >$thp_file
echo "Description=Disable Transparent Huge Pages (THP)" >>$thp_file
echo "DefaultDependencies=no" >>$thp_file
echo "After=local-fs.target" >>$thp_file
echo "Before=apparmor.service" >>$thp_file
echo "" >>$thp_file
echo "[Service]" >>$thp_file
echo "Type=oneshot" >>$thp_file
echo "ExecStart=/bin/sh -c 'echo never > /sys/kernel/mm/transparent_hugepage/enabled && echo never > /sys/kernel/mm/transparent_hugepage/defrag'" >>$thp_file
echo "" >>$thp_file
echo "[Install]" >>$thp_file
echo "WantedBy=multi-user.target" >>$thp_file
chmod 664 $thp_file
systemctl daemon-reload
systemctl enable disable-thp
systemctl start disable-thp
}
# 安装 kubeadm、kubelet 和 kubectl
install_kubernetes() {
echo "正在安装 kubeadm、kubelet 和 kubectl(版本:$kubernetes_version)..."
if [[ $os = "ubuntu" ]]; then
apt update
apt install -y apt-transport-https ca-certificates curl bridge-utils
modprobe br_netfilter # 加载所需的内核模块
curl -fsSL $apt_repository/doc/apt-key.gpg | apt-key add -
echo "deb $apt_repository kubernetes-xenial main" | tee /etc/apt/sources.list.d/kubernetes.list
apt update
apt install -y kubeadm=$kubernetes_version-00 kubelet=$kubernetes_version-00 kubectl=$kubernetes_version-00
elif [[ $os = "centos" || $os = "amazon_linux" ]]; then
cat <<EOF >/etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=${yum_repository}/yum/repos/kubernetes-el7-x86_64/
enabled=1
gpgcheck=0
repo_gpgcheck=0
gpgkey=${yum_repository}/yum/doc/yum-key.gpg
${yum_repository}/yum/doc/rpm-package-key.gpg
EOF
yum --debuglevel=1 install -y kubeadm-$kubernetes_version kubelet-$kubernetes_version kubectl-$kubernetes_version
systemctl enable kubelet
modprobe br_netfilter # 加载所需的内核模块
echo 1 > /proc/sys/net/bridge/bridge-nf-call-iptables
echo "添加bash-completion 自动补全"
yum install bash-completion -y
source /usr/share/bash-completion/bash_completion
source <(kubectl completion bash)
echo "source <(kubectl completion bash)" >>~/.bashrc
fi
}
# 安装 Containerd
install_containerd() {
echo "正在安装 Containerd..."
if [[ $os = "centos" || $os = "amazon_linux" ]]; then
yum install yum-utils -y
yum-config-manager --add-repo https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
yum --debuglevel=1 install -y containerd
elif [[ $os = "ubuntu" ]]; then
apt install -y containerd
fi
mkdir -p /etc/containerd
# 生成默认配置
containerd config default >/etc/containerd/config.toml
# 配置 systemd cgroup 驱动程序
sed -i 's#SystemdCgroup = false#SystemdCgroup = true#' /etc/containerd/config.toml
sed -i "s#registry.k8s.io#${docker_image_repository}#" /etc/containerd/config.toml
systemctl restart containerd
systemctl enable containerd
}
# 执行 kubeadm init 并复制 kubeconfig 文件
initialize_kubernetes_cluster() {
if command -v kubeadm &>/dev/null; then
kubeadm reset -f
else
echo "kubeadm 未找到,无法执行重置操作。请手动重置 Kubernetes。"
exit 1
fi
echo "正在执行 kubeadm init..."
kubeadm init --kubernetes-version=v${kubernetes_version}
--image-repository=${docker_image_repository}
--service-cidr=10.96.0.0/16
--pod-network-cidr=10.244.0.0/16
-v=5
# --kubernetes-version 指定要安装的Kubernetes版本
# --image-repository=registry.k8s.io 容器镜像仓库默认地址
# --service-cidr Kubernetes Service的IP地址范围
# --pod-network-cidr Kubernetes Pod的IP地址范围
# --control-plane-endpoint=test-k8s-lb.opsbase.cn:6443 控制平面终结点地址,用于在高可用集群中指定负载均衡器的地址。
echo "已成功执行 kubeadm init。"
# ctr 查看镜像list
ctr image ls
echo "正在复制 kubeconfig 文件..."
mkdir -p $HOME/.kube
cp /etc/kubernetes/admin.conf $HOME/.kube/config
chown $(id -u):$(id -g) $HOME/.kube/config
echo "kubeconfig 文件已复制到 $HOME/.kube/config。"
}
# 安装网络组件(Flannel)
install_network_plugin_flannel() {
echo "正在安装 Flannel 网络组件..."
echo $flannel
kubectl apply -f $flannel
}
# 安装网络组件(Calico)
install_network_plugin_calico() {
echo "正在安装 Calico 网络组件..."
kubectl create -f $calico
}
# 主函数
main() {
select_country
get_os_info
check_root_user
check_kubernetes_installed
disable_firewall
disable_swap
disable_transparent_hugepage
install_kubernetes
install_containerd
optimize_kernel
if [[ "$node_type" = "master" ]]; then
initialize_kubernetes_cluster
install_network_plugin_flannel
# 如果想使用 Calico 网络组件,注释掉上面的 "flannel" 函数,然后取消"calico" 行的注释
# install_network_plugin_calico
else
echo "slave节点,跳过集群初始化操作。"
fi
}
# 主函数
main
导入kubeconfig
代码语言:javascript复制# kubectl配置导入
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
# 查询集群信息
kubectl get cs
kubectl get node
Worker 节点
bash k8s.sh
代码语言:javascript复制#woker节点执行;加入到k8s集群
kubeadm join 192.168.0.212:6443 --token 2b1a67.5gbmuzdckkfnw9c6 --discovery-token-ca-cert-hash sha256:a9d0ebeccaxxxxxxxxxxxx
kubeadm更新证书
代码语言:javascript复制# 查看当前证书有效期
cd /etc/kubernetes/pki; for i in $(ls *.crt); do echo "===== $i ====="; openssl x509 -in $i -text -noout | grep -A 3 'Validity' ; done
# 更新证书有效期为10年
cd /etc/kubernetes/pki
mkdir backup_key; cp -rp ./* backup_key/
git clone https://github.com/yuyicai/update-kube-cert.git
cd update-kube-cert/
bash update-kubeadm-cert.sh all
代码语言:javascript复制## 重新查询证书有效期
===== apiserver.crt =====
Validity
Not Before: Aug 2 06:35:59 2023 GMT
Not After : Jul 30 06:35:59 2033 GMT
Subject: CN=kube-apiserver
===== apiserver-etcd-client.crt =====
Validity
Not Before: Aug 2 06:35:59 2023 GMT
Not After : Jul 30 06:35:59 2033 GMT
Subject: O=system:masters, CN=kube-apiserver-etcd-client
===== apiserver-kubelet-client.crt =====
Validity
Not Before: Aug 2 06:35:59 2023 GMT
Not After : Jul 30 06:35:59 2033 GMT
Subject: O=system:masters, CN=kube-apiserver-kubelet-client
===== ca.crt =====
Validity
Not Before: Aug 2 06:24:14 2023 GMT
Not After : Jul 30 06:24:14 2033 GMT
Subject: CN=kubernetes
===== front-proxy-ca.crt =====
Validity
Not Before: Aug 2 06:24:14 2023 GMT
Not After : Jul 30 06:24:14 2033 GMT
Subject: CN=front-proxy-ca
===== front-proxy-client.crt =====
Validity
Not Before: Aug 2 06:35:59 2023 GMT
Not After : Jul 30 06:35:59 2033 GMT
Subject: CN=front-proxy-client
集群设置
生成token给其他节点加入
代码语言:javascript复制# 主节点可重新生成 token , 配置为永久token可改ttl为0
kubeadm token create --print-join-command --ttl 0
master可调度pod (重要)
代码语言:javascript复制# 设置master可调度pod
kubectl get node
kubectl describe node logan | grep -i taint
kubectl taint nodes logan node-role.kubernetes.io/control-plane:NoSchedule-
# 禁止节点调度pod
kubectl taint nodes logan node-role.kubernetes.io/control-plane:NoSchedule
创建pod
代码语言:javascript复制# 提前拉取镜像(可选)
$ctr image pull docker.io/library/nginx:alpine
# 查看当前node上的镜像
$ ctr i ls
# 部署nginx,副本数为2
$ kubectl create deployment nginx --image nginx --port 80 --replicas=2
$ kubectl get po -owide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-55f598f8d-gfcgz 1/1 Running 0 70s 10.244.0.4 ubuntu-1 <none> <none>
nginx-55f598f8d-xn9jt 1/1 Running 0 70s 10.244.1.2 centos-1 <none> <none>
获取ingress-nginx版本
Welcome - NGINX Ingress Controller
helm安装 (测试建议买外网主机,防止无法镜像拉取不到)
代码语言:javascript复制# 安装,更多版本参考: https://github.com/helm/helm/releases
wget https://get.helm.sh/helm-v3.9.4-linux-amd64.tar.gz
tar -zxf helm-*-linux-amd64.tar.gz
cp linux-amd64/helm /usr/local/bin/
# 添加helm repo
helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo add harbor https://helm.goharbor.io
helm repo add bitnami https://charts.bitnami.com/bitnami
helm repo add Microsoft http://mirror.azure.cn/kubernetes/charts
helm repo update
helm repo list
# 搜索镜像
helm search repo node-exporter
helm search repo prometheus-operator
# 安装ingress
helm install ingress-nginx ingress-nginx/ingress-nginx
--namespace kube-system
# 安装监控
kubectl create ns monitor
helm -n monitor install node-exporter bitnami/node-exporter
#
报错问题
Centos7 worker节点加入 ubuntu k8s-master报错:
代码语言:javascript复制Normal Scheduled 18m default-scheduler Successfully assigned kube-system/kube-proxy-7tttz to test-1
Warning FailedCreatePodSandBox 3m29s (x71 over 18m) kubelet Failed to create pod sandbox: open /run/systemd/resolve/resolv.conf: no such file or directory
代码语言:javascript复制# 解决方案
mkdir -p /run/systemd/resolve/
sudo ln -sf /etc/resolv.conf /run/systemd/resolve/resolv.conf
# master
kubectl -n kube-system delete po kube-proxy-7tttz