k8s集群之-搭建

Kubernetes官网

1.前置要求

3台Linux 搭建1个master节点和2个node节点
每台机器 2 GB 或更多的运行内存(少了可能起不来)
每台机器2CPU或以上
集群中的所有机器的网络彼此均能相互(能ping通)连接(公网和内网都可以)
节点之中不可以有重复的主机名、MAC 地址或 product_uuid
开启机器上的某些端口
禁用交换分区

2.设置linux环境

关闭防火墙(开发时生产环境请自定义规则)
```
systemctl stop firewalld
systemctl disable firewalld
```

关闭selinux

#全局禁用
sed -i 's/enforcing/disabled/' /etc/selinux/config
#禁用当前会话
setenforce 0

关闭swap 内存交换

#关闭当前
swapoff -a
#全局禁用 
vim /etc/fstab
#将带有swap的注释掉 wq保存
#/dev/mapper/centos-swap swap    swap    defaults        0 0
#验证 swap必须全为0  -m:兆单位 -g:吉单位
free -m

设置主机名不能是localhost

#查看主机名
hostname
#设置主机名
hostnamectl set-hostname 新主机名

添加主机名与IP的对应关系

vim /etc/hosts

172.16.10.67 k8s-node1
172.16.10.68 k8s-node2
172.16.10.69 k8s-node3

3.安装

卸载之前的docker(所有节点)

sudo yum remove docker \
docker-client \
docker-client-latest \
docker-common \
docker-latest \
docker-latest-logrotate \
docker-logrotate \
docker-engine

安装docker的前置依赖(所有节点)

yum install -y yum-utils device-mapper-persistent-data lvm2

设置docker的yun源(所有节点)

sudo yum-config-manager \
--add-repo \
http://download.docker.com/linux/centos/docker-ce.repo

安装docker以及docker-ci(所有节点)

sudo yum install -y docker-ce docker-ce-di containerd.io

配置docker加速(所有节点)

sudo mkdir -p /etc/docker
sudo tee /etc/docker/daemon.json <<-'EOF'
{
  "registry-mirrors": ["https://akqgo94r.mirror.aliyuncs.com"]
}
EOF
sudo systemctl daemon-reload
sudo systemctl restart docker

设置开机启动(所有节点)
```
systemctl enable docker
```

添加阿里yum源(所有节点)

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=0
repo_gpgcheck=0
gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg
htsts://mirrors.aliyun.com/kubemnetes/yum/doc/rpm-package-key.gpg
EOF

安装kubeadm、kubelet、kubectl(所有节点)

#查看yum源  可以检测上一步配置的正确性
yum list|grep kube
#安装
yum install -y kubelet-1.17.3 kubeadm-1.17.3 kubectl-1.17.3
#启动kubelet和设置开机启动
systemctl enable kubelet
systemctl start kubelet

设置kubeadm、kubelet、kubectl镜像(只在master执行)

添加master_images.sh文件赋予可执行权限

#!/bin/bash

images=(
	kube-apiserver:v1.17.3
    	kube-proxy:v1.17.3
	kube-controller-manager:v1.17.3
	kube-scheduler:v1.17.3
	coredns:1.6.5
	etcd:3.4.3-0
    pause:3.1
)

for imageName in ${images[@]} ; do
    docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/$imageName
done

执行master_images.sh (只在master执行)

./master_images.sh

初始化master apiserver-advertise-address地址是你自己master的地址(只在master执行)
```
kubeadm init \
--apiserver-advertise-address=172.16.10.67 \
--image-repository registry.cn-hangzhou.aliyuncs.com/google_containers \
--kubernetes-version v1.17.3 \
--service-cidr=10.96.0.0/16 \
--pod-network-cidr=10.244.0.0/16
```
- 错误解决
  
  [preflight] Running pre-flight checks [WARNING IsDockerSystemdCheck]: detected "cgroupfs" as the Docker cgroup driver. The recommended driver is "systemd". Please follow the guide at https://kubernetes.io/docs/setup/cri/ error execution phase preflight: [preflight] Some fatal errors occurred: [ERROR FileContent--proc-sys-net-bridge-bridge-nf-call-iptables]: /proc/sys/net/bridge/bridge-nf-call-iptables contents are not set to 1 [preflight] If you know what you are doing, you can make a check non-fatal with–ignore-preflight-errors=…To see the stack trace of this error execute with --v=5 or higher
```
echo "1" >/proc/sys/net/bridge/bridge-nf-call-iptables
```

开始使用之前先按照提示执行几个命令(只在master执行)

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

安装网络插件(flannel)(只在master执行)
- 获取kube-flannel.yml 文件
```
wget http://oss.tianch.xyz/k8s/kube-flannel.yml
```
- apply配置文件
```
kubectl apply -f kube-flannel.yml
```
- 查看pods 都要Running才行
```
kubectl get pods --all-namespaces
```
查看主节点状态
```
kubectl get nodes
```

master节点Ready了让其它节点加入

kubeadm join 172.16.10.67:6443 --token idjrsn.y0g0ji2sbde1ev3r \
    --discovery-token-ca-cert-hash sha256:4204390176617c66a8b40f665a63bfc520a0ae031181006fbfd8b2c182653f6b

token过期解决 ttl 0为永不过期

kubeadm token create --ttl 0 --print-join-command

监控pod进度

watch kubectl get pod -n kube-system -o wide

全部Ready就安装完成了