调度主机：192.168.0.85；192.168.0.52
realserver:192.168.0.96;192.168.0.66

192.168.0.52:
到http://www.keepalived.org/download.html下载源码keepalived，我用的是keepalived-1.2.5.tar.gz
[root@desktop52 ~]# yum install kernel-devel openssl-devel libnl-devel popt-devel gcc make
如果不装kernel-devel可以编译成功，但安装后只有调度，没有lvs支持，因为他是调用内核的lvs，编译需要kernel-devel
解压，编译，如果要用snmp和debug也可已先安装再编译
configure后，可以看到Extra Lib:-lpot -lssl -lcrypto -lnl
默认安装目录为/usr/local/etc下，我们需要把相应文件靠到系统目录下（也可已在安装的时候指定安装目录为/etc）
[root@desktop52 etc]# pwd
/usr/local/etc
[root@desktop52 etc]# cp rc.d/init.d/keepalived /etc/init.d/
[root@desktop52 etc]# cp sysconfig/keepalived
keepalived/ rc.d/       sysconfig/ 
[root@desktop52 etc]# cp sysconfig/keepalived /etc/sysconfig/
[root@desktop52 keepalived]# ln -s /usr/local/sbin/keepalived /usr/sbin/ <===因为脚本里面的运行keepalived时只在/usr/sbin里面找
配置keepalived.conf
[root@desktop52 keepalived]# cat keepalived.conf
! Configuration File for keepalived
global_defs {
   notification_email {
     XXXX@gmail.com <===管理员邮箱
   }
   notification_email_from keepalived@desktop52 <===通知邮箱from
   smtp_server 127.0.0.1  <===使用本机转发 email
   smtp_connect_timeout 30
   router_id LVS_DEVEL <== load balancer 的标识 ID,用于 email 警报
}

vrrp_instance VI_1 {
    state MASTER <===备机改为 BACKUP,此状态是由 priority 的值来决定的,当前
priority 的值小于备机的值,那么将会失去 MASTER 状态
    interface eth0 <==#HA 监测网络接口
    virtual_router_id 57 <==主、备机的 virtual_router_id 必须相同,arbitary unique number 0..255
    priority 50
    advert_int 1 <===主备之间的通告间隔秒数
    authentication { <===主备切换时的验证
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {<====HA 虚拟 ip,可加多个
        192.168.0.136
    }
}

virtual_server 192.168.0.136 80 {
    delay_loop 6  <===每隔 6 秒查询 realserver状态
    lb_algo rr  <===lvs 调度算法,这里使用轮叫
    lb_kind DR
#    nat_mask 255.255.255.0
#    persistence_timeout 50 <===同一 IP 的连接 60 秒内被分配到同一台 realserver
    protocol TCP <===同一 IP 的连接 60 秒内被分配到同一台 realserver

    real_server 192.168.0.96 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3  <====故障重试秒数
            delay_before_retry 3 <===重试延迟
        }
    }
    real_server 192.168.0.66 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
}
备用机上该两个地方：
state，priority，而  notification_email_from 可改可不该
[root@desktop52 keepalived]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
启动时可以监控/var/log/messages日志文件,可以看到Activating healthchecker ...:80和Transition to MASTER STATE(备用机为BACKUP STATE)

realserver:（两台电脑都一样）
[root@desktop66 ~]# ifconfig lo:0 192.168.0.136 netmask 255.255.255.255 up <===lo上是默认是nonarp的所以不用更改内核,如果是别的端口，要注意该下面的内核配置：
net.ipv4.conf.all.arp_ignore=1
net.ipv4.conf.lo.arp_ignore=1
net.ipv4.conf.all.arp_announce=2
net.ipv4.conf.lo.arp_announce=2

[root@desktop66 ~]#  route add -host 192.168.0.136 dev lo:0
[root@desktop66 ~]# yum install -y httpd
[root@desktop66 ~]# vim /var/www/html/index.html
[root@desktop66 ~]# /etc/init.d/httpd start

[root@desktop85 ~]# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000
    link/ether 52:54:00:6c:e6:09 brd ff:ff:ff:ff:ff:ff
    inet 192.168.0.85/24 brd 192.168.0.255 scope global eth0
    inet 192.168.0.136/32 scope global eth0
    inet6 fe80::5054:ff:fe6c:e609/64 scope link
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast state DOWN qlen 1000
    link/ether 52:54:00:8d:82:cc brd ff:ff:ff:ff:ff:ff
看到我们的eth0中我们配置的136虚拟IP
注：这个我们只能用ip addr show查看调度器在哪台上运行，不想piranha和heartbeat可以ifconfig查看虚拟IP