keepalived高可用服务搭建——单实例

准备四台机器

hostname	IP	说明
lb01	10.0.0.30	keepalived主服务器(nginx主负载均衡)
lb02	10.0.0.60	Keepalived备服务器(Nginx辅负载均衡)
www01	10.0.0.40	www01
www02	10.0.0.50	www02

安装keepalived(lb01，lb02)

[root@lb01 ~]# yum install keepalived -y
[root@lb01 ~]# rpm -qa keepalived
keepalived-1.3.5-19.el7.x86_64

[root@lb02 ~]# yum install -y keepalived
[root@lb02 ~]# rpm -qa keepalived
keepalived-1.3.5-19.el7.x86_64

配置主keepalived服务器 lb01-master

1.备份旧配置文件，定义新配置文件
[root@lb01 keepalived]# cp keepalived.conf{,.bak} 
[root@lb01 keepalived]# ls
keepalived.conf  keepalived.conf.bak
2.根据自己机器修改配置
vrrp_instance VI_1 {
    state MASTER
    interface ens33
    virtual_router_id 51
    priority 150
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
#定义虚拟IP，也就是VIP，提供给用户的高可用地址，绑定网络接口ens33，别名ens33：1，主备节点要相同
    virtual_ipaddress {
        192.168.17.100/24 dev ens33 label ens33:1
    }
}
[root@lb01 keepalived]# systemctl restart keepalived
[root@lb01 keepalived]# ps -ef |grep keepalived
root       1299      1  0 10:38 ?        00:00:00 /usr/sbin/keepalived -D
root       1300   1299  0 10:38 ?        00:00:00 /usr/sbin/keepalived -D
root       1301   1299  0 10:38 ?        00:00:00 /usr/sbin/keepalived -D
root       1303   1200  0 10:38 pts/0    00:00:00 grep --color=auto keepalived
##检查是否生成新的VIP地址192.168.17.100
[root@lb01 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.30/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet 192.168.17.100/24 scope global ens33:1
    inet6 fe80::c537:24bd:f23f:695e/64 scope link noprefixroute 

配置keepalived备服务器lb02-BACKUP

注意lb02和lb01的参数区别

vrrp_instance VI_1 {
    state BACKUP
    interface ens33
    virtual_router_id 51　　#虚拟路由IP跟主节点一致
    priority 100　　　　#优先级备用节点一般比主节点低50
    advert_int 1　　　　#MASTER和BACKUP之间的通信检查时间，默认是1秒
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.17.100/24 dev ens33 label ens33:1
    }
}
[root@lb02 keepalived]# systemctl restart keepalived
[root@lb02 keepalived]# ps -ef |grep keepalived
root       1285      1  0 10:49 ?        00:00:00 /usr/sbin/keepalived -D
root       1286   1285  0 10:49 ?        00:00:00 /usr/sbin/keepalived -D
root       1287   1285  0 10:49 ?        00:00:00 /usr/sbin/keepalived -D
root       1293   1171  0 10:51 pts/0    00:00:00 grep --color=auto keepalived
[root@lb02 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.60/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet6 fe80::8d10:27e0:1c5b:e2a1/64 scope link tentative noprefixroute dadfailed 
    inet6 fe80::b5c6:7506:252b:95e/64 scope link noprefixroute

检查配置结果，是否出现虚拟IP（VIP）

• 正确的结果应该是没有出现VIP
• 因为lb02位BACKUP，当主节点正常的时候，它不会接管VIP
• 若是lb02看到了VIP，表示配置出错了！

如果出现lb01、lb02抢夺VIP资源，排查故障思路

• 主备服务器是否能够通信，可以ping检测
• 是否没有正确配置防火墙，或是忘记关闭防火墙
• 两台服务器的keepalived.conf是否写错（例如虚拟路由ID是否不一致）

高可用主备服务切换实践

停止主服务器的keepalived或者直接关机

[root@lb01 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.30/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet 192.168.17.100/24 scope global ens33:1
    inet6 fe80::c537:24bd:f23f:695e/64 scope link noprefixroute 
[root@lb01 keepalived]# systemctl stop keepalived
[root@lb01 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.30/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet6 fe80::c537:24bd:f23f:695e/64 scope link noprefixroute

可以看到，关闭keepalived服务后，VIP也消失了 

检查BACKUP机器状态

[root@lb02 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.60/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet 192.168.17.100/24 scope global ens33:1
    inet6 fe80::8d10:27e0:1c5b:e2a1/64 scope link tentative noprefixroute dadfailed 
    inet6 fe80::b5c6:7506:252b:95e/64 scope link noprefixroute 

可以看到备节点lb02此时已经接管了VIP，并且这个接管期间，备节点还会主动发送ARP广播，让局域网内所有的客户端更新本地ARP缓存表，以便访问新接管VIP服务的节点。

恢复主节点，查看VIP动态

[root@lb01 keepalived]# systemctl restart keepalived
[root@lb01 keepalived]# ip addr | grep inet
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
    inet 10.0.0.30/24 brd 10.0.0.255 scope global noprefixroute ens33
    inet 192.168.17.100/24 scope global ens33:1
    inet6 fe80::c537:24bd:f23f:695e/64 scope link noprefixroute

 此时会发现，再次启动keepalived主服务，MASTER很快会接管VIP，备服务器正确应该是释放了VIP，此时用户的请求也就再次转移到新机器了。

明确单实例主备模式配置文件区别

区别参数	MASTER参数	BACKUP参数
router_id	lb01	lb02
state	MASTER	BACKUP
priority	150	100