k8s集群 双主双从高可用架构
*可以看出master的状态是未就绪(NotReady**),之所以是这种状态是因为还缺少一个附件**flannel**,没有网络各Pod是无法通信的。下面这三个镜像就是master管理节点上前面初始化之前要加载的其中的三个 没有可以从master上直接打包成tar包cp过来或者直接下载。以上是主节点的安装部署,然后是node1 node2节点的安装,和加入集群,这里注意,验证,node01节点已
主机清单
master: 192.168.27.10 docker kubelet kubectl kubeadm keepalived
master1: 192.168.27.20 docker k8s kubelet kubectl kubeadm keepalived
node1: 192.168.27.30 docker k8s kubelet kubeadm
node2: 192.168.27.40 docker k8s kubelet kubeadm
更改主机名
hostnamectl set-hostname master
hostnamectl set-hostname master1
hostnamectl set-hostname node01
hostnamectl set-hostname node02
添加对应域名解析
cat >> /etc/hosts << EOF
192.168.27.10 master
192.168.27.20 master1
192.168.27.30 node1
192.168.27.40 node2
EOF
防火墙,SElinux
systemctl stop firewalld
systemctl disable firewalld
setenforce 0
sed -i 's/SELINUX=enforcing/SELINUX=disabled/' /etc/selinux/config
时间同步
timedatectl set-timezone Asia/Shanghai;timedatectl set-local-rtc 0
禁用swap(关闭虚拟内存,储存在硬盘的,影响读写效率)
swapoff -a && sed -i '/swap/s/^/#/' /etc/fstab
free -h 查看禁用效果
配置集群无密登录 master到master1 node1,node2无密码
ssh-keygen -t rsa
ssh-copy-id root@master1
ssh-copy-id root@node1
ssh-copy-id root@node2
优化内核参数
[root@master ~]# modprobe br_netfilter
cat > /etc/sysctl.d/kubernetes.conf << EOF
net.bridge.bridge-nf-call-iptables = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.ipv4.ip_forward = 1
net.ipv4.tcp_tw_recycle = 0
vm.swappiness = 0
vm.overcommit_memory = 1
vm.panic_on_oom = 0
fs.inotify.max_user_instances = 8192
fs.inotify.max_user_watches = 1048576
fs.file-max = 52706963
fs.nr_open = 52706963
net.ipv6.conf.all.disable_ipv6 = 1
net.netfilter.nf_conntrack_max = 2310720
EOF
sysctl -p /etc/sysctl.d/kubernetes.conf
#如果这个命令显示某个目录没有 就重启一下docker
[root@master ~]# scp /etc/sysctl.d/kubernetes.conf node1:/etc/sysctl.d/
[root@master ~]# scp /etc/sysctl.d/kubernetes.conf node2:/etc/sysctl.d/
[root@master ~]# scp /etc/sysctl.d/kubernetes.conf master1:/etc/sysctl.d/
[root@node01 ~]# sysctl -p /etc/sysctl.d/kubernetes.conf
[root@node02 ~]# sysctl -p /etc/sysctl.d/kubernetes.conf
[root@master1 ~]# sysctl -p /etc/sysctl.d/kubernetes.conf
所有服务器安装docker
配个源
阿里云 ***
$ sudo yum-config-manager \
--add-repo \
http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
安装前提
yum install -y yum-utils device-mapper-persistent-data lvm2
安装19.03.10版本docker
yum -y install docker-ce-19.03.10 docker-ce-cli-19.03.10 containerd.io
systemctl start docker
更改docker默认Cgroup驱动
在/etc/docker/daemon.json文件中,添加一句话即可,当然这个和我们的设置加速器写在一起了。
mkdir /etc/docker
cat >> /etc/docker/daemon.json << EOF
{
"registry-mirrors": ["https://1dmptu91.mirror.aliyuncs.com"],
"exec-opts": ["native.cgroupdriver=systemd"]
}
EOF
systemctl daemon-reload
systemctl restart docker
所有master节点安装haproxy和keepalived
yum -y install haproxy keepalived
#主master节点
mv /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak
vim /etc/keepalived/keepalived.conf
global_defs {
router_id lb1
#路由id号,和主服务器必须不同(删除vrrp_strict行)
}
vrrp_script chk_haproxy {
script "/data/check_haproxy.sh"
interval 2
weight 2
}
vrrp_instance VI_1 {
state MASTER
#状态:BACKUP备 MASTER主
interface ens33
virtual_router_id 51
priority 100
#优先级:备比主要小
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
192.168.27.254
#虚拟路由ip,公共ip
}
}
#备主master1
mv /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak
把主master的拷贝过来 改个优先级和状态为 BACKUP
修改haproxy配置文件
两台主节点 都改一样的 最后的ip换成自己的两个主节点ip 后面端口号别动
mv /etc/haproxy/haproxy.cfg /etc/haproxy/haproxy.cfg.bak
vim /etc/haproxy/haproxy.cfg
global
log /dev/log local0
log /dev/log local1 notice
pidfile /var/run/haproxy.pid
chroot /var/lib/haproxy
stats socket /var/run/haproxy-admin.sock mode 660 level admin
stats timeout 30s
user haproxy
group haproxy
daemon
nbproc 1
defaults
log global
timeout connect 5000
timeout client 10m
timeout server 10m
listen admin_stats
bind 0.0.0.0:10080
mode http
log 127.0.0.1 local0 err
stats refresh 30s
stats uri /status
stats realm welcome login\ Haproxy
stats auth admin:123456
stats hide-version
stats admin if TRUE
listen kube-master
bind 0.0.0.0:8443
mode tcp
option tcplog
balance source
server master 192.168.27.10:6443 check inter 2000 fall 2 rise 2 weight 1
server master1 192.168.27.20:6443 check inter 2000 fall 2 rise 2 weight 1
都启动
systemctl start haproxy
systemctl enable haproxy
systemctl start keepalived
systemctl enable keepalived
看下漂移ip
ip a show dev ens33
是否在第一台上
两台主安装kubelet kubectl kubeadm
两台从安装kubelet kubeadm
所有版本都必须相同
yum install kubelet-1.18.0 kubeadm-1.18.0 kubectl-1.18.0 -y --nogpgcheck
都启动
systemctl start kubelet
systemctl enable kubelet
主master:
在初始化之前需要先把一些下载好镜像文件 提前导入好
导入本地镜像文件
docker load -i 跟名字
共是8个包,我都提前下载直接随便拉某个目录下加载就行
coredns.tar
etcd.tar
kube-apiserver.tar
kube-controller-manager.tar
kube-proxy.tar
pause.tar
flannel.tar
kube-scheduler.tar
kubeadm初始化
kubeadm init --kubernetes-version=v1.18.0 --image-repository registry.aliyuncs.com/google_containers --apiserver-advertise-address 192.168.27.254:8443 --pod-network-cidr=10.244.0.0/16 --service-cidr=10.96.0.0/12
成功后提示:
下面这个是两个主都要创的 立刻去创
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
下面这两个结果做保存
下面这个是备主连主的命令
kubeadm join 192.168.27.254:8443 --token hps85k.h4zgc1vrcq0hw9bd \
--discovery-token-ca-cert-hash sha256:8057206b9e00740b2dc52b6957b516460ecdf6168885b3502ed058349d48015d \
--control-plane --certificate-key 3f4f9625c46eaa17a5ce99f0d64436d87d270f2719374b67036e6b8f828d178f
下面这个是node节点连主的命令
kubeadm join 192.168.27.254:8443 --token hps85k.h4zgc1vrcq0hw9bd \
--discovery-token-ca-cert-hash sha256:8057206b9e00740b2dc52b6957b516460ecdf6168885b3502ed058349d48015d
重启主master
systemctl restart kubelet
kubectl get nodes
NAME STATUS ROLES AGE VERSION
master NotReady master 5m3s v1.18.0
**可以看出master的状态是未就绪(NotReady**),之所以是这种状态是因为还缺少一个附件**flannel**,没有网络各Pod是无法通信的
//添加网络组件(flannel),组件flannel可以通过https://github.com/coreos/flannel中获取
kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
再次查看集群节点状态
[root@master ~]# kubectl get nodes
NAME STATUS ROLES AGE VERSION
master Ready master 13m v1.18.0
设置kubectl命令行工具自动补全功能(备主master1 也要下载)
yum install -y bash-completion
source /usr/share/bash-completion/bash_completion
source <(kubectl completion bash)
echo "source <(kubectl completion bash)" >> ~/.bashrc
以上是主节点的安装部署,然后是node1 node2节点的安装,和加入集群,这里注意,验证,node01节点已经准备好了相关的镜像。
两台node节点操作一样
下面这条命令就是前面初始化做保存的 在上直接复制过来确定
[root@node01 ~]# kubeadm join 192.168.27.10:6443 --token njus35.kw3hxkys3urmnuob --discovery-token-ca-cert-hash sha256:05761b73b571c18eebd6972fb70323cd3c4d8e0aa7514efa2680411310424184
node节点需要几个镜像,可以手动先下载一下,或者从master节点同步
这个可以先查看下都有没有
docker images (查看镜像)
下面这三个镜像就是master管理节点上前面初始化之前要加载的其中的三个 没有可以从master上直接打包成tar包cp过来或者直接下载
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/kube-proxy:v1.18.0
docker pull quay.io/coreos/flannel:v0.15.1
docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/pause:3.2
保存镜像为tar包
docker save -o nginx.tar nginx:latest
备主master1 :
下面这条命令就是前面初始化做保存的 直接复制过来
[root@master1 ~]# kubeadm join 192.168.27.254:8443 --token hps85k.h4zgc1vrcq0hw9bd \
--discovery-token-ca-cert-hash sha256:8057206b9e00740b2dc52b6957b516460ecdf6168885b3502ed058349d48015d \
--control-plane --certificate-key 3f4f9625c46eaa17a5ce99f0d64436d87d270f2719374b67036e6b8f828d178f
[root@master ~]# kubectl get nodes (没出错的情况下 等node01 Ready 需要多等一会)
NAME STATUS ROLES AGE VERSION
master Ready master 18m v1.18.0
master1 Ready master 18m v1.18.0
node1 Ready <none> 68s v1.18.0
node2 Ready <none> 68s v1.18.0
> PS: 确保所有pod都是running状态。(没出错的情况下 等所有running 需要多等一会)
//在master验证k8s组件 如下所示代表安装组件完整且是running状态
[root@master ~]# kubectl get pod -n kube-system
去除Master节点污点,使其可以分配Pod资源
##下面的命令在备主master1执行即可
kubectl taint nodes --all node-role.kubernetes.io/master-
验证高可用
关掉主master
systemctl stop kubelet
正常 漂移地址就会飘到备主上
ip a show dev ens33
##主master关机后正常情况下VIP地址会跳到备主master1上
##从上面信息可以看到VIP地址飘逸成功,查看各个pod是否正常。
[root@master1 ~]# kubectl get pod --all-namespaces
可以看的到pod都是Running状态。
##查看各个节点状态:
[root@master1 ~]# kubectl get nodes
NAME STATUS ROLES AGE VERSION
master NotReady control-plane,master 2d1h v1.18.0
master1 Ready control-plane,master 2d1h v1.18.0
node1 Ready control-plane,master 2d1h v1.18.0
node2 Ready <none> 2d1h v1.18.0
#可以看的出来master1状态为NotReady
更多推荐
所有评论(0)