目录
| 主机名 | IP |
| nginx01 | 11.0.1.31 |
| nginx01 | 11.0.1.31 |
| lvs01 | 11.0.1.33 |
| lvs02 | 11.0.1.34 |
| VIP1 | 11.0.1.29 |
| VIP2 | 11.0.1.30 |
一、原理
lvs+keepalived+nginx主备模式下，lvs永远只有一台在工作，资源利用率并不高，双主模式可以解决这一问题。
确保真实IP11.0.1.31、11.0.1.32访问正常
二、真实服务器nginx配置
nginx01、nginx02 执行real_server.sh脚本
新建real_server.sh脚本并执行，脚本内容为lo网卡新增两个虚拟接口，添加linux内核路由，关闭arp功能。
[root@nginx01 ~]# cat /etc/keepalived/real_server.sh
ifconfig lo:0 11.0.1.29 netmask 255.255.255.255 broadcast 11.0.1.29
route add -host 11.0.1.29 dev lo:0
ifconfig lo:1 11.0.1.30 netmask 255.255.255.255 broadcast 11.0.1.30
route add -host 11.0.1.30 dev lo:1
echo "1" >/proc/sys/net/ipv4/conf/lo/arp_ignore
echo "2" >/proc/sys/net/ipv4/conf/lo/arp_announce
echo "1" >/proc/sys/net/ipv4/conf/all/arp_ignore
echo "2" >/proc/sys/net/ipv4/conf/all/arp_announce
[root@nginx01 ~]# chmod +x /etc/keepalived/real_server.sh
[root@nginx01 ~]# sh /etc/keepalived/real_server.sh查看nginx01网卡信息
查看nginx02网卡信息
三、lvs的keepalived配置
3.1 配置文件
注意有两个VRRP实例，两个VIP，一个VRRP实例对应一个VIP
| VRRP 1（VIP 11.0.1.29） | 11.0.1.33 master 11.0.1.34 slave |
| VRRP 2（VIP11.0.1.30） | 11.0.1.33 slave 11.0.1.34 master |
lvs01-33
[root@lvs01-33 ~]# cat /etc/keepalived/keepalived.conf
vrrp_instance VI_1 {
state MASTER
interface ens33
virtual_router_id 31
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass abcd
}
virtual_ipaddress {
11.0.1.29
}
}
vrrp_instance VI_2 {
state BACKUP
interface ens33
virtual_router_id 41
priority 99
advert_int 1
authentication {
auth_type PASS
auth_pass abcd
}
virtual_ipaddress {
11.0.1.30
}
}
virtual_server 11.0.1.29 80 {
delay_loop 6
lb_algo rr
lb_kind DR
protocol TCP
real_server 11.0.1.31 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
real_server 11.0.1.32 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
}
virtual_server 11.0.1.30 80 {
delay_loop 6
lb_algo rr
lb_kind DR
protocol TCP
real_server 11.0.1.31 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
real_server 11.0.1.32 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
}lvs02-34
[root@lvs02-34 ~]# cat /etc/keepalived/keepalived.conf
vrrp_instance VI_1 {
state BACKUP
interface ens33
virtual_router_id 31
priority 99
advert_int 1
authentication {
auth_type PASS
auth_pass abcd
}
virtual_ipaddress {
11.0.1.29
}
}
vrrp_instance VI_2 {
state MASTER
interface ens33
virtual_router_id 41
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass abcd
}
virtual_ipaddress {
11.0.1.30
}
}
virtual_server 11.0.1.29 80 {
delay_loop 6
lb_algo rr
lb_kind DR
protocol TCP
real_server 11.0.1.31 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
real_server 11.0.1.32 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
}
virtual_server 11.0.1.30 80 {
delay_loop 6
lb_algo rr
lb_kind DR
protocol TCP
real_server 11.0.1.31 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
real_server 11.0.1.32 80 {
weight 1
inhibit_on_failure
TCP_CHECK {
connect_timeout 3
nb_get_retry 2
delay_before_retry 1
connect_port 80
}
}
}3.2 开启keepalived服务
查看lvs01网卡信息
查看lvs02网卡信息
四、测试
4.1 测试访问VIP
4.2 模拟lvs01宕机
模拟lvs01宕机，关闭lvs1的keepalived
lvs02网卡信息
访问测试，两个虚拟VIP仍然可以正常使用，依旧可以正常访问nginx01、nginx02,，轮询方式仍然正常工作，注意每次刷新前都须有清除cookie记录。
重启lvs01的keepalived，虚拟VIP11.0.1.29重新漂移回lvs01（11.0.1.33）
模拟lvs02宕机同理，不再重复实验。
参考文档：
（2）LVS+Keepalived+nginx高可用负载均衡架构原理及配置_51CTO博客_keepalived nginx负载均衡配置
