Nginx+Keepalived

1.  规划环境：

2.  配置Nginx负载均衡，这里通过子配置文件进行配置：

        1>.  Nginx_Master服务器Nginx是通过源码安装，配置文件在/usr/local/nginx/conf/nginx.conf

        2>.  Nginx_Backup服务器Nginx使用过yum源安装，配置文件在/etc/nginx/nginx.conf

##Nginx_Master服务器Nginx配置
[root@centosnode1 ~]# vim /usr/local/nginx/conf.d/lb2.conf 
 
upstream websrvs {
        server 192.168.188.227:80 weight=1;
        server 192.168.188.228:80 weight=1;
}
 
server {
        location / {
                proxy_pass http://websrvs;
                index index.html;
        }
}

##Nginx_Backup服务器Nginx配置
[root@centosnode2 ~]# vim /etc/nginx/conf.d/lb2.conf
 
upstream websrvs {
        server 192.168.188.227:80 weight=1;
        server 192.168.188.228:80 weight=1;
}
 
server {
        location / {
                proxy_pass http://websrvs;
                index index.html;
        }
}

 3.  启动Nginx服务，测试负载均衡是否成功：

##启动主,从Nginx服务
[root@centosnode1 ~]# systemctl start nginx.service
[root@centosnode2 ~]# systemctl start nginx.service
 
##设置Web1服务器的简单web页面
[root@centosnode3 ~]# echo "this is page ip : 192.168.188.227" > /var/www/html/index.html
 
##启动Web1服务器的httpd服务
[root@centosnode3 ~]# systemctl start httpd
 
##设置Web2服务器的简单web页面
[root@centosnode4 ~]# echo "this is page ip : 192.168.188.228" > /var/www/html/index.html
 
##启动Web2服务器的httpd服务
[root@centosnode4 ~]# systemctl start httpd
 
##测试Nginx服务负载均衡
[root@centosnode5 ~]# curl 192.168.188.223
this is page ip : 192.168.188.227 
[root@centosnode5 ~]# curl 192.168.188.223
this is page ip : 192.168.188.228

4.  配置Nginx_Master服务器Keepalived服务：

[root@centosnode1 ~]# vim /etc/keepalived/keepalived.conf
 
! Configuration File for keepalived
 
global_defs {
   notification_email {
     acassen@firewall.loc
     failover@firewall.loc
     sysadmin@firewall.loc
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 192.168.200.1
   smtp_connect_timeout 30
   router_id NODE1_MASTER
}
 
vrrp_script chk_nginx {
        script "killall -0 nginx"  ##判断nginx服务是否运行，若不运行，则返回状态码非0
        interval 2
}
 
vrrp_instance VI_1 {
    state MASTER
    interface ens33
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
 
    track_script {  ##调用vrrp_script中定义的脚本
        chk_nginx
    }
 
    virtual_ipaddress {
        192.168.188.100
    }
}

5.  配置Nginx_Backup服务器Keepalived服务：

[root@centosnode2 ~]# vim /etc/keepalived/keepalived.conf
 
! Configuration File for keepalived
 
global_defs {
   notification_email {
     acassen@firewall.loc
     failover@firewall.loc
     sysadmin@firewall.loc
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 192.168.200.1
   smtp_connect_timeout 30
   router_id NODE2_BACKUP
}
 
vrrp_script chk_nginx {
        script "killall -0 nginx"  ##判断nginx服务是否运行，若不运行，则返回状态码非0
        interval 2
}
 
vrrp_instance VI_1 {
    state BACKUP
    interface ens33
    virtual_router_id 51
    priority 80
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
 
    track_script {  ##调用vrrp_script中定义的脚本
        chk_nginx
    }
 
    virtual_ipaddress {
        192.168.188.100
    }
}

6.  启动Keepalived服务：

[root@centosnode1 ~]# systemctl start keepalived.service
 
[root@centosnode2 ~]# systemctl start keepalived.service

7.  测试Nginx负载均衡是否正常，访问VIP：

##测试Nginx服务负载均衡
[root@centosnode5 ~]# curl 192.168.188.100
this is page ip : 192.168.188.227 
[root@centosnode5 ~]# curl 192.168.188.100
this is page ip : 192.168.188.228

8.  测试故障检查是否正常，停掉一台web后端的web服务，访问VIP是否可以正常访问：

##停掉centosnode3的httpd服务
[root@centosnode3 ~]# systemctl stop httpd
 
##访问vip
[root@centosnode5 ~]# for ((i=1;i<=5;i++))
> do
> curl 192.168.188.100
> done
this is page ip : 192.168.188.228
this is page ip : 192.168.188.228
this is page ip : 192.168.188.228
this is page ip : 192.168.188.228
this is page ip : 192.168.188.228

由此可见，当centosnode3节点的httpd故障时，会自动将其从集群中去除；而且访问vip可以正常访问。当centosnode3节点的httpd恢复时，会自动将其加入到集群中。

9.  测试Keepalived故障转换是否正常，是否会发生ip飘移：

由此可见当前vip地址在centosnode1主机上，然后模拟服务器故障，停掉Nginx服务。

[root@centosnode1 ~]# systemctl stop nginx.service

 

再次查看vip，发现vip飘移到了centosnode2主机上。可以将centosnode1主机的Nginx服务再次开启，会发现vip又会飘移到了centosnode1主机上，因为keepalived默认配置抢占主服务器，优先级高的服务器，将抢占到vip。

[root@centosnode1 ~]# systemctl start nginx.service

##监控日志文件
tail -f 日志文件目录