环境:centos 6.4 64bit
应用:nginx
目的:keepalived可以让两台服务器处于主备关系,如果主的挂了,备的取得VIP(或者互为主备等关系,文字游戏不纠结),
以实现服务器的高可用。
关于恢复:
如果设置keepalived为主备模式,那么当主服务器恢复,VIP重新回到主服务器上,
那么就产生了多次切换的问题,所以这里我不采用主备,而是采用备备的模式。
但是备备模式,两台机的优先级不一样,还是会竞争,所以需要在优先级高的服务器上设置nopreempt(具体见下面的配置)。
两台服务器,备备模式,都运行着nginx:
nginx-1: 192.168.1.136
nginx-1: 192.168.1.150
VIP(虚拟IP):192.168.1.176
1
|
yum install -y gcc gcc-c++ popt-devel openssl openssl-devel libssl-dev libnl-devel popt-devel |
安装keepalived
1
2
3
4
5
6
|
wget http: //www.keepalived.org/software/keepalived-1.2.13.tar.gz
tar zxvf keepalived- 1.2 . 13 .tar.gz
cd keepalived- 1.2 . 13
./configure --prefix=/usr/local/keepalived make make install |
keepalived开机启动,两个脚本(开机启动可以参考我另外的文章。)
vim setkeep_startup.sh
1
2
3
4
5
6
7
8
|
#!/bin/bash mkdir /opt/sta chmod a+x /opt/sta/* echo -e "start on runlevel 2\nstart on runlevel 3\nstart on runlevel 4\nstart on runlevel 5\nrespawn\nexec /opt/sta/keep_check.sh" > /etc/init/keep.conf
initctl reload-configuration initctl list initctl start keep |
vim keep_check.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
|
#!/bin/bash #check keepalived while [ 1 ]
do Run=$(ps aux|grep keep|grep -v "grep" |wc -l)
echo Run=$Run if [ 0 == $Run ]
then /usr/local/keepalived/sbin/keepalived echo "start keepalived"
else echo "keepalived is runing"
fi sleep 3
done |
nginx开机启动
同上,改改就行。如果连改都懒得改,请自行右上角。
nginx检测脚本(两台服务器都需要),如果发现Nginx进程不在了,则杀死keepalived进程:
1
2
3
4
5
6
7
8
|
#!/bin/bash run=`ps -C nginx --no-header |wc -l` if [ $run -eq 0 ];then
killall keepalived
echo "kill keepalived" >> /opt/keep.log
else echo "nginx alive" >> /opt/keep.log
fi |
nginx-1配置,路径:/usr/local/keepalived/etc/keepalived/keepalived.conf
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
|
vrrp_script check_nginx { script "/opt/sta/nginx_alive_check.sh" #nginx监控脚本
interval 2 #执行监控时间间隔
weight 2 #脚本结果导致的优先级变更: 2 表示优先级+ 2 ;- 2 表示优先级- 2
} global_defs { notification_email {
root@localhost
}
notification_email_from root@local host
smtp_server localhost
smtp_connect_timeout 30
router_id NodeA
} vrrp_instance VI_1 { state BACKUP #两台机都设置为BACKUP
interface eth0 #绑定虚拟IP的网络接口
virtual_router_id 51 #VRRP组名,两个节点的设置必须一样,以指明各个节点属于同一VRRP组
priority 100 #竞争时节点的优先级( 1 - 254 之间)
nopreempt #在优先级高的机器上设置,防止恢复时重新竞争
advert_int 3 #组播信息发送间隔,两个节点设置必须一样
authentication { #设置验证信息,两个节点必须一致
auth_type PASS
auth_pass 1234
}
track_script { check_nginx #监控服务定义
}
virtual_ipaddress { #虚拟IP, 两个节点设置一样
192.168 . 1.176 / 24
}
} |
nginx-2配置,路径:/usr/local/keepalived/etc/keepalived/keepalived.conf
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
|
vrrp_script check_nginx { script "/opt/sta/nginx_alive_check.sh" #nginx监控脚本
interval 2 #执行监控时间间隔
weight 2 #脚本结果导致的优先级变更: 2 表示优先级+ 2 ;- 2 表示优先级- 2
} global_defs { notification_email {
root@localhost
}
notification_email_from root@local host
smtp_server localhost
smtp_connect_timeout 30
router_id NodeA
} vrrp_instance VI_1 { state BACKUP #两台机都设置为BACKUP
interface eth1 #绑定虚拟IP的网络接口
virtual_router_id 51 #VRRP组名,两个节点的设置必须一样,以指明各个节点属于同一VRRP组
priority 99 #主节点的优先级( 1 - 254 之间),备用节点必须比主节点优先级低
advert_int 3 #组播信息发送间隔,两个节点设置必须一样
authentication { #设置验证信息,两个节点必须一致
auth_type PASS
auth_pass 1234
}
track_script { check_nginx #监控服务定义
}
virtual_ipaddress { #虚拟IP, 两个节点设置一样
192.168 . 1.176 / 24
}
} |
启动keepalived
cd /usr/local/keepalived/sbin/
./keepalived -f /usr/local/keepalived/etc/keepalived/keepalived.conf
-f为指定配置文件路径
启动之后在/var/log/messages中可以看到相关信息输出
输出ip addr可以查看网卡上绑定的ip,看看是不是VIP已经绑定到某一台机器的网卡了。
如果发现两台机都绑了,那么就是两台机无法正常通信,检查防火墙。
关于防火墙,据说这样设置可以让keepalived正常工作:
到/etc/sysconfig/iptables中添加一行
-A INPUT -m state --state NEW -m tcp -p tcp -d 224.0.0.0/8 -j ACCEPT
-A INPUT -i eth0 -p vrrp -j ACCEPT
但是我发现不设置,但是stop然后start,也可以正常工作。目前未解。