CLUSTER集群Keepalived热备+LVS服务

keepalived

         通过在两台主、备调度器上实现Keepalived高可用解决调度器单点失败问题

         主、备调度器上配置LVS，正常情况下主调度器工作

         主调度器异常时，Keepalived启用备用调度器，保证了调度器的高可用

         调度器保证了后台服务的高可用

调度器安装keepalived

# yum install  -y  gcc gcc-c++  kernel-devel  openssl-devel popt-devel

# tar  xvzf  keepalived-1.2.7.tar.gz  -C  /usr/src/

# cd  /usr/src/keepalived-1.2.7/

# ./configure  --sysconf=/etc  --with-kernel-dir=/usr/src/kernels/2.6.32-358.el6.x86_64/ && make  &&  make install

#ln  -s  /usr/local/sbin/keepalived /sbin/

# chkconfig  keepalived  on

二、Keepalived配置说明

设置报警邮件

global_defs {

 notification_email {

   [email protected]

  }

 notification_email_from ka@localhost

 smtp_server 192.168.20.1

 smtp_connect_timeout 30

 router_id  LVS_devel

}

 

VRRP实例设置

vrrp_instance VI_1 {

 state MASTER     //主服务器为MASTER，辅助为SLAVE

 interface eth0

 virtual_router_id 51

 priority 100     //主服务器优先级要比辅助的高

 advert_int 1

 authentication {

   auth_type pass

   auth_pass forlvs   //主辅服务器密码必须一致

  }

 virtual_ipaddress { 192.168.20.100  }

}

 

virtual_server 192.168.20.100 80 {   //设置虚拟IP为192.168.20.100

 delay_loop 6

 lb_algo rr          //设置LVS调度算法为RR

 lb_kind DR       //设置LVS的模式为DR

 persistence_timeout 50

 protocol TCP

 real_server 192.168.20.150 80 {

   weight 3          //设置权重为3

   TCP_CHECK {

         connect_timeout3

         nb_get_retry3

         delay_before_retry3

    }

  }

#Real Server参见VLS-DR部分的配置