image.png

上面的拓?fù)鋱D，LB1和LB2是兩個(gè)Nginx反向代理，后端RS1和RS2是兩個(gè)Web服務(wù)器。
接下來(lái)我們通過(guò)keepalived實(shí)現(xiàn)高可用的nginx反向代理集群。

keepalived的實(shí)現(xiàn)本質(zhì)是：
當(dāng)LB1和LB2都正常運(yùn)行時(shí)，把VIP 172.16.80.201分配給高優(yōu)先級(jí)的LB1。用戶訪問(wèn)時(shí)，就由LB1提供反向代理。
當(dāng)LB1出現(xiàn)故障時(shí)，keepalived就會(huì)把VIP漂移到LB2，由LB2提供反向代理功能。從而保證了業(yè)務(wù)的正常運(yùn)行。

RS1和RS2配置

1、安裝Nginx

[root@RS1 ~]#yum install nginx -y
[root@RS2 ~]#yum install nginx -y

2、配置訪問(wèn)頁(yè)面，并啟動(dòng)RS1和RS2的Nginx服務(wù)

[root@RS1 ~]#vim /usr/share/nginx/html/index.html 
RS1 172.16.80.100
[root@RS2 ~]#vim /usr/share/nginx/html/index.html 
RS2 172.16.80.102

客戶端訪問(wèn)RS1和RS2效果如下：
web頁(yè)面如下：

image.png

LB1和LB2的配置

1、LB1和LB2的Nginx負(fù)載均衡配置

[root@LB1 nginx]#vim /etc/nginx/nginx.conf
worker_processes  4;
events {
    worker_connections  1024;
}
http {
    include       mime.types;
    default_type  application/octet-stream;
    server_tokens off;
    sendfile        on;
    keepalive_timeout  65;
    upstream www.server.pools {
        server 172.16.80.100:80 weight=1;
        server 172.16.80.102:80 weight=1;
    }
    server {
        listen       80;
        server_name  www.keepalived.com;
        location / {
        proxy_pass http://www.server.pools;
        proxy_set_header Host $host;
        proxy_set_header X-Forwardered-For $remote_addr;
        }

        }
}

LB2的Nginx配置與LB1一樣。配置后重啟nginx服務(wù)?？蛻舳朔謩e訪問(wèn)172.16.80.101和172.16.80.103，可以看到LB實(shí)現(xiàn)了負(fù)責(zé)均衡的反向代理：

image.png

2、LB1和LB2通過(guò)keepalived實(shí)現(xiàn)高可用
安裝keepalived

[root@LB1 ~]#yum install keepalived ipvsadm  -y
[root@LB2 ~]#yum install keepalived ipvsadm  -y

配置VRRP實(shí)例，LB1為master，LB22為backup。
VS1配置：

[root@LB1 ~]#vim /etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {
   notification_email {
   root@localhost
}
   notification_email_from keadmin@localhost
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id CentOS7A.luo.com
   vrrp_mcast_group4 224.0.0.22
}

vrrp_instance VI_1 {
    state MASTER
    interface ens33
    virtual_router_id 51
    priority 150
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass haha
    }
    virtual_ipaddress {
        172.16.80.201/16
    }

}

LB2配置

[root@LB2 ~]#vim /etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {                                            <==全局配置段開(kāi)始
   notification_email {                                  
   root@localhost                                        <==故障郵件的收件人
}                                                    
   notification_email_from keadmin@localhost             <==故障郵件的發(fā)件人
   smtp_server 127.0.0.1                                 <==故障郵件的發(fā)件服務(wù)器
   smtp_connect_timeout 30
   router_id CentOS7B.luo.com                            <==路由器ID
   vrrp_mcast_group4 224.0.0.22                          <==組播地址
}                                                        <==全局配置段結(jié)束 

vrrp_instance VI_1 {                                     <==實(shí)例名字為VI_1，備節(jié)點(diǎn)的實(shí)例名字要和主節(jié)點(diǎn)的相同
    state BACKUP                                         <==狀態(tài)為backup
    interface ens33                                      <==通信的接口
    virtual_router_id 51                                 <==實(shí)例ID
    priority 100                                         <==優(yōu)先級(jí)
    advert_int 1                                         <==通信檢查時(shí)間間隔
    authentication {
        auth_type PASS                                   <==認(rèn)證類(lèi)型
        auth_pass haha                                   <==認(rèn)證密碼
    }
    virtual_ipaddress {
        172.16.80.201/16                                 <==虛擬IP
    }
}

配置完后，啟動(dòng)VS1和VS2的keepalived

image.png

image.png

上圖可以看到，LB1成功拿到VIP 172.16.80.201。

驗(yàn)證

image.png

客戶端訪問(wèn)http://172.16.80.201時(shí)，被輪詢調(diào)度到RS1和RS2。
當(dāng)我把LB1關(guān)機(jī)后，VIP就會(huì)被keepalived分配到LB2。這時(shí)客戶端再次訪問(wèn)http://172.16.80.201，仍然被輪詢調(diào)度到后端RS服務(wù)器，用戶是無(wú)法感知到LB1的故障的。