K8S多节点部署

一、k8s多节点部署（接上篇博客）

1、环境准备

服务器	ip	组件
k8s集群master01	192.168.142.3	kube-apiserver、kube-controller-manager、kube-scheduler、etcd
k8s集群node1	192.168.142.4	kubelet、kube-proxy、docker、flannel
k8s集群node2	192.168.142.5	kubelet、kube-proxy、docker、flannel
k8s集群master02	192.168.142.6	kube-apiserver、kube-controller-manager、kube-scheduler、etcd
负载均衡器1	192.168.142.10
负载均衡器2	192.168.142.20
VIP	192.168.142.100

2、从master01节点上拷贝证书文件、配置文件和服务管理文件到master02

1 2 3

scp -r /opt/etcd/ root@192.168.142.6:/opt/ scp -r /opt/kubernetes/ root@192.168.142.6:/opt scp /usr/lib/systemd/system/{kube-apiserver,kube-controller-manager,kube-scheduler}.service root@192.168.142.6:/usr/lib/systemd/system/

 

3、修改配置文件kube-apiserver中的IP

1 2 3 4 5 6 7

vim /opt/kubernetes/cfg/kube-apiserver KUBE_APISERVER_OPTS="--logtostderr=true \ --v=4 \ --etcd-servers=https://192.168.142.3:2379,https://192.168.142.4:2379,https://192.168.142.5:2379 \ --bind-address=192.168.142.6 \              #修改master02的ip --secure-port=6443 \ --advertise-address=192.168.142.6 \         #修改master02的ip

 

4、在master02节点上启动各服务并设置开机自启

1 2 3

systemctl enable --now kube-apiserver.service systemctl enable --now kube-controller-manager.service systemctl enable --now kube-scheduler.service

 

5、查看node节点状态

1 2 3 4

ln -s /opt/kubernetes/bin/* /usr/local/bin/ kubectl get nodes kubectl get nodes -o wide           #-o=wide：输出额外信息；对于Pod，将输出Pod所在的Node名 //此时在master02节点查到的node节点状态仅是从etcd查询到的信息，而此时node节点实际上并未与master02节点建立通信连接，因此需要使用一个VIP把node节点与master节点都关联起来

二、负载均衡部署（nginx实现负载均衡，keepalived实现双机热备）（192.168.142.10 192.168.142.20）

1、配置nginx的官方在线yum源，配置本地nginx的yum源

1 2 3 4 5 6 7 8

cat > /etc/yum.repos.d/nginx.repo << 'EOF' [nginx] name=nginx repo baseurl=http://nginx.org/packages/centos/7/$basearch/ gpgcheck=0 EOF   yum install nginx -y

2、修改nginx配置文件，配置四层反向代理负载均衡

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

##指定k8s群集2台master的节点ip和6443端口   vim /etc/nginx/nginx.conf events {     worker_connections  1024; }   #添加 stream {     log_format  main  '$remote_addr $upstream_addr - [$time_local] $status $upstream_bytes_sent';           access_log  /var/log/nginx/k8s-access.log  main;       upstream k8s-apiserver {         server 192.168.142.3:6443;         server 192.168.142.6:6443;     }     server {         listen 6443;         proxy_pass k8s-apiserver;     } }   http { ......

3、检查配置文件语法并启动nginx服务

1 2 3 4 5 6

1.检查配置文件语法 nginx -t     2.启动nginx服务，查看已监听6443端口 systemctl enable --now nginx netstat -natp | grep nginx

4、部署keepalived服务

1	yum install keepalived -y

 

5、修改keepalived配置文件（额外编写健康检查脚本）

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

vim /etc/keepalived/keepalived.conf ! Configuration File for keepalived   global_defs {    # 接收邮件地址    notification_email {      acassen@firewall.loc      failover@firewall.loc      sysadmin@firewall.loc    }    # 邮件发送地址    notification_email_from Alexandre.Cassen@firewall.loc    smtp_server 127.0.0.1    smtp_connect_timeout 30    router_id NGINX_MASTER   #lb01节点的为NGINX_MASTER，lb02节点的为 NGINX_BACKUP }   #添加一个周期性执行的脚本 vrrp_script check_nginx {     script "/etc/nginx/check_nginx.sh"  #指定检查nginx存活的脚本路径 }   vrrp_instance VI_1 {     state MASTER            #lb01节点的为 MASTER，lb02节点的为 BACKUP     interface ens33         #指定网卡名称 ens33     virtual_router_id 51    #指定vrid，两个节点要一致     priority 100            #lb01节点的为 100，lb02节点的为 90     advert_int 1     authentication {         auth_type PASS         auth_pass 1111     }     virtual_ipaddress {         192.168.142.100      #指定 VIP     }     track_script {         check_nginx         #指定vrrp_script配置的脚本     } } ========================================================== ##创建nginx状态检查脚本 vim /etc/nginx/check_nginx.sh #!/bin/bash #egrep -cv "grep|$$" 用于过滤掉包含grep 或者 $$ 表示的当前Shell进程ID count=$(ps -ef | grep nginx | egrep -cv "grep|$$")   if [ "$count" -eq 0 ];then     systemctl stop keepalived fi     chmod +x /etc/nginx/check_nginx.sh

 

6、启动keepalived服务（一定要先启动了nginx服务，再启动keepalived服务）

1 2 3

systemctl enable --now nginx systemctl enable --now keepalived ip a    #查看VIP是否生成

三、修改node节点上的kubeconfig配置文件

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

//修改bootstrap.kubeconfig,kubelet.kubeconfig配置文件为VIP   cd /opt/kubernetes/cfg/ vim bootstrap.kubeconfig server: https://192.168.80.100:6443                         vim kubelet.kubeconfig server: https://192.168.80.100:6443                           vim kube-proxy.kubeconfig server: https://192.168.80.100:6443   //重启kubelet和kube-proxy服务 systemctl restart kubelet.service systemctl restart kube-proxy.service

 

 

四、在master01节点上操作

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

1.测试创建pod kubectl create deployment nginx-test --image=nginx   2.查看Pod的状态信息 kubectl get pod kubectl get pods -o wide   3.在对应网段的node节点上操作，可以直接使用浏览器或者curl命令访问 curl 172.17.45.2   4.这时在master01节点上查看nginx日志，发现没有权限查看 kubectl logs nginx-test-7d965f56df-v42rp   Error from server (Forbidden): Forbidden (user=system:anonymous, verb=get, resource=nodes, subresource=proxy) (nginx-test-7d965f56df-v42rp)   5.在master01节点上，将cluster-admin角色授予用户system:anonymous 命令：kubectl create clusterrolebinding cluster-system-anonymous --clusterrole=cluster-admin --user=system:anonymous clusterrolebinding.rbac.authorization.k8s.io/cluster-system-anonymous created   6.再次查看nginx日志 kubectl logs nginx-test-7d965f56df-v42rp   7.在master01和master02上分别创建pod来测试负载均衡 master01：已创建上面的pod master02：kubectl create deployment nginx-testaa --image=nginx