Keepalived + Haproxy for high-availability load balancing

Keepalived is a two-machine hot-standby (HA) software, often and haproxy combined to do a warm standby + load balancing, to achieve high availability

1.Experimental Environment

master:10.20.2.233

backup:10.20.2.235

Virtual ip:10.20.2.238

2.keepalived Installation

1) install software dependency package via Yum

Yum install-y wget make gcc openssl-devel popt-devel

2) keepalived source package Download Installation

: http://www.keepalived.org/software/keepalived-1.2.23.tar.gz

TAR-ZXF keepalived-1.2.23.tar.gz-c/usr/srccd/sru/src/keepalived-1.2.23/./configure--prefix=/usr/local/ Keepalived/makemake Install

3) service mode after successful installation

cp/usr/local/keepalived/sbin/keepalived/usr/sbin/cp/usr/local/keepalived/etc/sysconfig/keepalived/etc/ sysconfig/cp/usr/local/keepalived/etc/rc.d/init.d/keepalived/etc/init.d/

3.Editing a configuration file

Mkdir-p/etc/keepalivedcp/usr/local/keepalived/etc/keepalived/keepalived.conf/etc/keepalived/

master configuration file are as follows:

! configuration file for keepalived global_defs {   router_id  HA_TEST}  #监测haproxy进程状态, Vrrp_script chk_haproxy {    script once every 2 seconds   "/etc/keepalived/check_haproxy.sh"     interval 2    weight  2} vrrp_instance VI_1 {    state MASTER #  from for backup From     priority 200 #  to 100    virtual_router_id 51   #路由ID and can be viewed via #tcpdump vrrp.     garp_master_delay 1  #主从切换时间 units in seconds.     interface eth0    advert_int 1     mcast_src_ip 10.20.2.233     #MASTER服务器IP, write ip     from server to server authentication {        auth_type pass         auth_pass 1111    }    virtual_ipaddress {         10.20.2.238     #虚拟IP     }     track_script {        chk_haproxy    # Monitor Haproxy process status     }}

backup configuration file are as follows:

! configuration file for keepalived global_defs {   router_id  HA_TEST}  #监测haproxy进程状态, Vrrp_script chk_haproxy {    script once every 2 seconds   "/etc/keepalived/check_haproxy.sh"     interval 2    weight  2} vrrp_instance VI_1 {    state BACKUP #  Master as Master     priority 100 #  Master for 101    virtual_router_id 51   #路由ID and can be viewed via #tcpdump vrrp.     garp_master_delay 1  #主从切换时间 units in seconds.     interface eth0    advert_int 1     mcast_src_ip 10.20.2.235     #BACKUP服务器IP, the master server writes the ip     of the primary server authentication {        auth_type pass         auth_pass 1111    }    virtual_ipaddress {         10.20.2.238     #虚拟IP     }     track_script {        chk_haproxy    # Monitor Haproxy process status     }}

3) haproxy detection Script (check_haproxy.sh) content is as follows

#!/bin/bash a= ' ps-c haproxy--no-header |wc-l ' If [$A-eq 0];then/etc/init.d/haproxyd startsleep 3 if [' Ps-c ha Proxy--no-header |wc-l '-eq 0];then/etc/init.d/keepalived stop fi fichmod 755/etc/keepalived/check_haproxy.sh #脚 You need to assign permissions after this writing is completed.

4) keepalived Start-Stop

Service keepalived start #停止service keepalived stop #启动service keepalived restart #重启

4.Test:

1) start the keepalived on master and backup

Service keepalived Start (keepalived starts at the same time as haproxy) [[email protected] keepalived]# service keepalived startstarting keepalived: [OK][[email protected] keepalived]# service Haproxyd statushaproxy (p ID 32077) is running!

Master binds the Vip,backup unbound VIP

[[Email protected] keepalived]# ip a1: lo: <loopback,up,lower_up> mtu  65536 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 forever2: eth0: <broadcast ,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state up qlen  1000    link/ether 00:50:56:85:46:b2 brd ff:ff:ff:ff:ff:ff     inet 10.20.2.233/24 brd 10.20.2.255 scope global eth0   &NBSP;&NBSP;INET&NBSP;10.20.2.238/32&NBSP;SCOPE&NBSP;GLOBAL&NBSP;ETH0&NBSP;&NBSP;&NBSP;&NBSP;INET6&NBSP;FE80: : 250:56ff:fe85:46b2/64 scope link        valid_lft forever preferred_lft  Forever[[email protected] keepalived]# ip a1: lo: <loopback,up,lower_up>  mtu 65536 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 forever2: eth0: < Broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state up qlen  1000    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff     inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0     inet6 fe80::250: 56ff:fe85:73ce/64 scope link        valid_lft forever  preferred_lft forever

2) when the haproxy,3 seconds are stopped, the keepalived will start

[[Email protected] keepalived]# service haproxyd stopstop HAProxy listenstop HAProxy process[[email protected] keepalived ]# service Haproxyd statushaproxy (PID 32368) is running!

3) Stop the keepalived of the Lord and take over the service immediately .

Backup's got a VIP on the set.

[[Email protected] keepalived]# ip a1: lo: <loopback,up,lower_up> mtu  65536 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 forever2: eth0: <broadcast ,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state up qlen  1000    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff     inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0   &NBSP;&NBSP;INET&NBSP;10.20.2.238/32&NBSP;SCOPE&NBSP;GLOBAL&NBSP;ETH0&NBSP;&NBSP;&NBSP;&NBSP;INET6&NBSP;FE80: : 250:56ff:fe85:73ce/64 scope link        valid_lft forever preferred_lft  Forever

start the Keepalived,master binding vip,backup release VIP

[[Email protected] keepalived]# service keepalived startstarting keepalived:                                          [  OK  ][[email protected] keepalived]# ip a1:  Lo: <loopback,up,lower_up> mtu 65536 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  Forever2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_ Fast state up&nbSp;qlen 1000    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff     inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0     inet 10.20.2.238/32 scope global eth0    inet6  fe80::250:56ff:fe85:73ce/64 scope link        valid_ Lft forever preferred_lft forever

This article is from the "11528244" blog, please be sure to keep this source http://11538244.blog.51cto.com/11528244/1851606

Keepalived + Haproxy for high-availability load balancing