Keepalived + Haproxy 实现高可用负载均衡

56341

keepalived是一个用于做双机热备（HA）的软件，常和haproxy联合起来做热备＋负载均衡，达到高可用

1. 实验环境Master:10.20.2.233
Backup:10.20.2.235
虚拟IP:10.20.2.238
2. Keepalived安装

1) 通过yum安装软件依赖包

1	yum install -y wget make gcc openssl-devel popt-devel

2) Keepalived源码包下载安装

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

1 2 3 4 5

tar -zxf keepalived-1.2.23.tar.gz -C /usr/src cd /sru/src/keepalived-1.2.23/ ./configure --prefix=/usr/local/keepalived/ Make make install

3) 安装成功后做成服务模式

1 2 3

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. 编辑配置文件

1 2	mkdir -p /etc/keepalived cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/

1) Master配置文件内容如下：

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

! Configuration File for keepalived    global_defs {    router_id HA_TEST }    #监测haproxy进程状态，每2秒执行一次 vrrp_script chk_haproxy {     script "/etc/keepalived/check_haproxy.sh"     interval 2     weight 2 }    vrrp_instance VI_1 {     state MASTER # 从为BACKUP     priority 200 # 从为100     virtual_router_id 51 #路由ID，可通过#tcpdump vrrp查看。     garp_master_delay 1 #主从切换时间，单位为秒。     interface eth0     advert_int 1     mcast_src_ip 10.20.2.233    #MASTER服务器IP,从服务器写从服务器的IP     authentication {         auth_type PASS         auth_pass 1111     }     virtual_ipaddress {         10.20.2.238    #虚拟IP     }     track_script {         chk_haproxy    #监测haproxy进程状态     } }

2) BACKUP配置文件内容如下：

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

! Configuration File for keepalived    global_defs {    router_id HA_TEST }    #监测haproxy进程状态，每2秒执行一次 vrrp_script chk_haproxy {     script "/etc/keepalived/check_haproxy.sh"     interval 2     weight 2 }    vrrp_instance VI_1 {     state BACKUP # 主为MASTER     priority 100 # 主为101     virtual_router_id 51 #路由ID，可通过#tcpdump vrrp查看。     garp_master_delay 1 #主从切换时间，单位为秒。     interface eth0     advert_int 1     mcast_src_ip 10.20.2.235    #BACKUP服务器IP,主服务器写主服务器的IP     authentication {         auth_type PASS         auth_pass 1111     }     virtual_ipaddress {         10.20.2.238    #虚拟IP     }     track_script {         chk_haproxy    #监测haproxy进程状态     } }

3) haproxy检测脚本（check_haproxy.sh）内容如下

1 2 3 4 5 6 7 8 9 10 11

#!/bin/bash   A=`ps -C haproxy --no-header |wc -l`   if [ $A -eq 0 ];then   /etc/init.d/haproxyd start sleep 3   if [ `ps -C haproxy --no-header |wc -l` -eq 0 ];then   /etc/init.d/keepalived stop   fi   Fi chmod 755 /etc/keepalived/check_haproxy.sh   #脚本编写完成后需要赋上权限

4) Keepalived的启停

1 2 3

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

4. 测试：

1) 启动MASTER和BACKUP上的keepalived

1 2 3 4 5

service keepalived start （keepalived启动的同时会将haproxy启动） [iyunv@test keepalived]# service keepalived start Starting keepalived:                                       [  OK  ] [iyunv@test keepalived]# service haproxyd status HAProxy(pid 32077) is running!

MASTER绑定了VIP,BACKUP未绑定VIP

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

[iyunv@test keepalived]# ip a 1: 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 forever 2: 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     inet 10.20.2.238/32 scope global eth0     inet6 fe80::250:56ff:fe85:46b2/64 scope link        valid_lft forever preferred_lft forever [iyunv@dengqp01 keepalived]# ip a 1: 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 forever 2: 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) 停止主上的haproxy，3秒后，keepalived会将其启动

1 2 3 4 5

[iyunv@test keepalived]# service haproxyd stop STOP HAProxy Listen STOP HAProxy process [iyunv@test keepalived]# service haproxyd status HAProxy(pid 32368) is running!

3) 停止主上的keepalived，备机马上接管服务

BACKUP上绑定了VIP

1 2 3 4 5 6 7 8 9 10 11 12

[iyunv@dengqp01 keepalived]# ip a 1: 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 forever 2: 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     inet 10.20.2.238/32 scope global eth0     inet6 fe80::250:56ff:fe85:73ce/64 scope link        valid_lft forever preferred_lft forever

4) 启动主上的keepalived，MASTER绑定VIP，BACKUP释放了VIP

1 2 3 4 5 6 7 8 9 10 11 12 13 14

[iyunv@test keepalived]# service keepalived start Starting keepalived:                                       [  OK  ] [iyunv@dengqp01 keepalived]# ip a 1: 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 forever 2: 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     inet 10.20.2.238/32 scope global eth0     inet6 fe80::250:56ff:fe85:73ce/64 scope link        valid_lft forever preferred_lft forever