Heartbeat+Haproxy实现负载均衡高可用

环境说明：

主机名	角色	IP地址
	heartbeat+haproxy	eth0: eth1:
	heartbeat+haproxy	eth0: eth1:
	web server 1	eth0:
	web server 2	eth0:

一、环境准备

在mylinux1和mylinux2上安装heartbeat，可以手动安装，当然也可以使用前面发布的一篇heartbeat脚本的博文，里面有heartbeat快速安装脚本。

然后在mylinux2上安装haproxy，并将mylinux1上的haproxy配置文件和启停脚本都拷贝到mylinux2上。

要保证mylinux1和mylinux2的heartbeat+haproxy环境一致。

[root@mylinux1 ~]# scp /etc//haproxy mylinux2:/etc//
root@mylinux2's password: 
Permission denied, please try again.
root@mylinux2's password: 
haproxy                                       100% 1567     /s   00:00    
[root@mylinux1 ~]# scp /usr/local/haproxy/conf/ mylinux2:/usr/local/haproxy/conf/
root@mylinux2's password: 
                                   100% 1194     /s   00:00   
[root@mylinux2 haproxy]# ll /etc//haproxy 
-rwxr-xr-x 1 root root 1567 Oct  1 02:11 /etc//haproxy
[root@mylinux2 haproxy]# ll /usr/local/haproxy/conf/ 
-rw-r--r-- 1 root root 1194 Oct  1 02:13 /usr/local/haproxy/conf/

二、修改heartbeat的资源文件

[root@mylinux1 ~]# tail -1 /etc//haresources 
 IPaddr::/24/eth0 haproxy
[root@mylinux2 haproxy]# tail -1 /etc//haresources 
 IPaddr::/24/eth0 haproxy

这里还是贴出两个heartbeat节点的配置，如下：

[root@mylinux1 ~]# egrep -v "#|^$" /etc//
debugfile /var/log/ha-debug
logfile/var/log/ha-log
logfacilitylocal1
keepalive 2
deadtime 30
warntime 10
initdead 120
udpport694
mcast eth1  694 1 0
auto_failback on


respawn hacluster /usr/lib64/heartbeat/ipfail
[root@mylinux2 haproxy]# egrep -v "#|^$" /etc// 
debugfile /var/log/ha-debug
logfile/var/log/ha-log
logfacilitylocal1
keepalive 2
deadtime 30
warntime 10
initdead 120
udpport694
mcast eth1  694 1 0
auto_failback on
node    
node    
respawn hacluster /usr/lib64/heartbeat/ipfail

三、启动heartbeat服务

1、首先在mylinux1和mylinux2上先关闭haproxy服务

[root@mylinux1 ~]# service haproxy stop
Stop haproxy successful.
[root@mylinux2 haproxy]# service haproxy stop
Haproxy is not running. Stop haproxy failed!

2、依次启动mylinux1和mylinux2上的heartbeat服务

[root@mylinux1 ~]# /etc//heartbeat start
Starting High-Availability services: INFO:  Resource is stopped
Done.
[root@mylinux2 haproxy]# /etc//heartbeat start
Starting High-Availability services: INFO:  Resource is stopped
Done.

四、测试haproxy的高可用

1、首先，检查mylinux1的状态

[root@mylinux1 ~]# ip a|grep 
    inet /24 brd  scope global secondary eth0
[root@mylinux1 ~]# service haproxy status
Haproxy (pid  3600) is running...

2、其次，检查mylinux2的状态

[root@mylinux2 haproxy]# ip a |grep 
[root@mylinux2 haproxy]# service haproxy status
Haproxy is stopped.

3、查看监控界面

五、测试故障切换

1、停止mylinux1上的heartbeat服务

[root@mylinux1 ~]# /etc//heartbeat stop
Stopping High-Availability services: Done.

2、检查mylinux1和mylinux2的状态

[root@mylinux1 ~]# ip a |grep 
[root@mylinux1 ~]# service haproxy status
Haproxy is stopped.
[root@mylinux2 haproxy]# ip a |grep 
    inet /24 brd  scope global secondary eth0
[root@mylinux2 haproxy]# service haproxy status
Haproxy (pid  3997) is running...

3、查看监控页面

4、查看mylinux2上的heartbeat日志

[root@mylinux2 ~]# >/var/log/ha-log 

[root@mylinux2 ~]# tail -f /var/log/ha-log 

Oct 01 02:44:31 heartbeat: [3534]: info: Received shutdown notice from ''.

Oct 01 02:44:31 heartbeat: [3534]: info: Resources being acquired from .

Oct 01 02:44:31 heartbeat: [3655]: info: acquire local HA resources (standby).

Oct 01 02:44:31 heartbeat: [3656]: info: No local resources [/usr/share/heartbeat/ResourceManager listkeys ] to acquire.

Oct 01 02:44:31 heartbeat: [3655]: info: local HA resource acquisition completed (standby).

Oct 01 02:44:31 heartbeat: [3534]: info: Standby resource acquisition done [all].

harc(default)[3681]:2016/10/01_02:44:31 info: Running /etcstatus status

mach_down(default)[3698]:2016/10/01_02:44:31 info: Taking over resource group IPaddr::/24/eth0

ResourceManager(default)[3725]:2016/10/01_02:44:31 info: Acquiring resource group: IPaddr::/24/eth0 haproxy

/usr/lib/ocf///heartbeat/IPaddr(IPaddr_)[3753]:       2016/10/01_02:44:31 INFO:  Resource is stopped

ResourceManager(default)[3725]:2016/10/01_02:44:31 info: Running /etc///IPaddr /24/eth0 start

IPaddr(IPaddr_)[3878]:2016/10/01_02:44:32 INFO: Adding inet address /24 with broadcast address to device eth0

IPaddr(IPaddr_)[3878]:2016/10/01_02:44:32 INFO: Bringing device eth0 up

IPaddr(IPaddr_)[3878]:2016/10/01_02:44:32 INFO: /usr/libexec/heartbeat/send_arp -i 200 -r 5 -p /var/run/resource-agents/send_arp- eth0 auto not_used not_used

/usr/lib/ocf///heartbeat/IPaddr(IPaddr_)[3852]:       2016/10/01_02:44:32 INFO:  Success

ResourceManager(default)[3725]:2016/10/01_02:44:32 info: Running /etc//haproxy  start

mach_down(default)[3698]:2016/10/01_02:44:32 info: /usr/share/heartbeat/mach_down: nice_failback: foreign resources acquired

mach_down(default)[3698]:2016/10/01_02:44:32 info: mach_down takeover complete for node .

Oct 01 02:44:32 heartbeat: [3534]: info: mach_down takeover complete.

Oct 01 02:45:02 heartbeat: [3534]: WARN: node : is dead

Oct 01 02:45:02 heartbeat: [3534]: info: Dead node gave up resources.

Oct 01 02:45:02 heartbeat: [3534]: info: Link :eth1 dead.

Oct 01 02:45:02 ipfail: [3543]: info: Status update: Node now has status dead

Oct 01 02:45:03 ipfail: [3543]: info: NS: We are dead. :<

Oct 01 02:45:03 ipfail: [3543]: info: Link Status update: Link /eth1 now has status dead

Oct 01 02:45:04 ipfail: [3543]: info: We are dead. :<

Oct 01 02:45:04 ipfail: [3543]: info: Asking other side for ping node count.