预先步骤
｜ 目标安装版本 kubernetes 1.24.3
服务器配置：
| |
CPU |
Memory |
Disk |
System |
Ip |
| k8s-node-1 |
2C |
8G |
100GB |
Centos 7.9 |
192.168.50.50 |
| k8s-node-1 |
2C |
8G |
100GB |
Centos 7.9 |
192.168.50.51 |
| k8s-node-1 |
2C |
8G |
100GB |
Centos 7.9 |
192.168.50.52 |
1 配置Linux，安装依赖
1.1 update yum package到最新
yum update -y
1.2 关闭Swap，k8s worker node 禁止使用swap
shell 执行
编辑 /etc/fstab, 注释swap配置
1
| # /dev/mapper/centos-swap swap swap defaults 0 0
|
1.3 关闭firewalld
1
2
| systemctl disable firewalld
systemctl stop firewalld
|
1.4 设置SELINUX
1
2
| setenforce 0
sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config
|
1.5 设置node hostname
1
| hostnamectl set-hostname k8s-node-1
|
1.6 开启ip_forward
1
| echo 1 > /proc/sys/net/ipv4/ip_forward
|
1.7 开启bridge-nf-call-iptables
1
2
3
| modprobe br_netfilter
echo 1 > /proc/sys/net/bridge/bridge-nf-call-ip6tables
echo 1 > /proc/sys/net/bridge/bridge-nf-call-iptables
|
1.8 /etc/hosts 配置，保证node 间可以通过hostname互相访问
vim /etc/hosts
1
2
3
| k8s-node-1 192.168.50.50
k8s-node-2 192.168.50.51
k8s-node-3 192.168.50.52
|
1.9 重启机器
2 运行时（Containerd）安装
因为K8s在1.24以后已经不再支持docker作为容器运行时，所以我们选择安装containerd作为容器运行时
2.1 通过yum安装Containerd
1
2
3
4
5
| yum install -y yum-utils device-mapper-persistent-data lvm2
yum-config-manager --add-repo http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
yum install containerd -y
systemctl enable containerd
systemctl restart containerd
|
2.2 修改containerd的配置
1
| containerd config default > /etc/containerd/config.toml
|
修改/etc/containerd/config.toml,需要修改两个地方
2.2.1 修改cgroup为systemd
SystemdCgroup = true
2.2.2 修改sandbox_image
sandbox_image = “registry.aliyuncs.com/google_containers/pause:3.7”
保存 config.toml后退出
2.3 重启containerd
1
2
| systemctl daemon-reload
systemctl restart containerd
|
3 安装k8s
3.1 配置kubernetes 源
1
2
3
4
5
6
7
8
9
| cat <<EOF | sudo tee /etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=http://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64
enabled=1
gpgcheck=0
repo_gpgcheck=0
gpgkey=http://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg http://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
EOF
|
3.2 安装kubeadm , kubelet, kubectl
1
| yum install -y kubelet-1.24.3 kubeadm-1.24.3 kubectl-1.24.3 --disableexcludes=kubernetes
|
3.3 配置kubeadm config yaml, 示例配置如下， 保存为kubeadmin-config.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
| apiVersion: kubeadm.k8s.io/v1beta3
bootstrapTokens:
kind: InitConfiguration
localAPIEndpoint:
advertiseAddress: 192.168.50.50 ## 替换为master ip
bindPort: 6443
nodeRegistration:
criSocket: unix:///var/run/containerd/containerd.sock ## containerd sock
imagePullPolicy: IfNotPresent
name: k8s-node-1
taints:
- effect: NoSchedule
key: node-role.kubernetes.io/master
---
apiServer:
timeoutForControlPlane: 4m0s
apiVersion: kubeadm.k8s.io/v1beta3
certificatesDir: /etc/kubernetes/pki
clusterName: nuc-k8s
controllerManager: {}
dns: {}
etcd:
local:
dataDir: /var/lib/etcd
imageRepository: registry.aliyuncs.com/google_containers
kind: ClusterConfiguration
kubernetesVersion: v1.24.3
networking:
dnsDomain: cluster.local
podSubnet: 10.244.0.0/16 # pod 子网，需要后续和cni 插件的subnet 保持一致
serviceSubnet: 10.96.0.0/12
scheduler: {}
---
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
cgroupDriver: systemd # cgroup 驱动，与containerd cgroup 驱动保持一致
|
3.4 安装前预先拉取镜像
1
| kubeadm config images pull --config kubeadm-config.yaml
|
拉取完成后就可以执行安装了
3.5 执行安装
1
| kubeadm init --config kubeadm-config.yaml
|
安装完成后 可以看到从日志中看到如下类似的信息
1
| kubeadm join 192.168.50.50:6443 --token xxxx --discovery-token-ca-cert-hash sha256:xxxxx
|
复制保存，在其他node 上执行上面的命令，就可以加入这个集群了
** 至此集群已经初步搭建完成 **
但是 如果你执行 kubectl get nodes 或者 kubectl get po -A, 你会发现 workernode 还处于 unready 状态， 很多pod 比如 kubedns 还处于pending 状态，那是因为我们还有最后一步要做，安装cni 插件, 在此之前，先配置kubeconfig
3.5.1 kubeconfig 配置
1
2
3
| mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
|
3.6 安装cni插件，以calico为例
3.6.1 下载calico yaml 配置
1
| curl https://docs.projectcalico.org/manifests/calico.yaml
|
yaml中有一个地方需要我们修改
1
2
| - name: CALICO_IPV4POOL_CIDR
value: "10.244.0.0/16" ## 此处需与kubeadm podsubnet配置保存一致
|
1
| kubectl apply -f calico.yaml
|
4 安装完成
