五．网页监控--mfscgiserv
mfscgiserv是用python编写的一个web服务工具，它的监听端口是9425，可以利用：/usr/local/mfs/sbin/mfscgiserv来启动，用户利用浏览器就可全面监控所有客户挂接，chunkserver及master server，客户端的各种操作等等，绝对是个好工具。通过浏览器可以查看：:9425
六．MFS客户端常用命令介绍
[root@mfs-client mfs]# ll /usr/local/mfs/bin/
total 556
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsappendchunks -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfscheckfile -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsdeleattr -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsdirinfo -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsfileinfo -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsfilerepair -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsgeteattr -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsgetgoal -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsgettrashtime -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsmakesnapshot -> mfstools
-rwxr-xr-x 1 root root 384447 Jun 5 01:09 mfsmount
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsrgetgoal -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsrgettrashtime -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsrsetgoal -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsrsettrashtime -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfsseteattr -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfssetgoal -> mfstools
lrwxrwxrwx 1 root root 8 Jun 5 01:09 mfssettrashtime -> mfstools
-rwxr-xr-x 1 root root 137 Jun 5 01:09 mfssnapshot
-rwxr-xr-x 1 root root 170799 Jun 5 01:09 mfstools
七．整个系统架构总结
1.heartbeat与drbd的集成
DRBD的主从的切换依赖与heartbeat的服务，所以集成的关键点在于定义Initdrbd的脚本文件的定义和heartbeat的服务定义文件/etc/ha.d/resource.d。如mfs-master Initdrbd 就是集成DRBD和heartbeat。
2.heartbeat与MFS的集成
与DRBD一样，mfsmaster的脚本文件定义与heartbeat的服务定义文件/etc/ha.d/resource.d中即可。
3.drbd 与mfs 的集成
在主服务器中,必须把MFS的安装文件全部安装到drbd的挂载目录中去，安装到块设备上去，所以在主从切换的时候，备份机器拿到安装环境和主的一样，那么通过同样的方式来启动MFS进程，达到切换切换启动MFS的目的。
八．测试
1.两台server上启动DRBD，利用cat /proc/drbd 查看启动情况和网络状态
主机器上显示Primary/Secondary 备份机器上显示Secondary/Primary，启动成功
主上查看：
[root@mfs-master /]# cat /proc/drbd
version: 8.3.13 (api:88/proto:86-96)
GIT-hash: 83ca112086600faacab2f157bc5a9324f7bd7f77 build by mockbuild@builder10.CentOS.org, 2012-05-07 11:56:36
1: cs:Connected ro:Primary/Secondary ds:UpToDate/UpToDate C r-----
ns:2377 nr:41 dw:2418 dr:694 al:4 bm:1 lo:0 pe:0 ua:0 ap:0 ep:1 wo:b oos:0
备上查看：
[root@mfs-bak ha.d]# cat /proc/drbd
version: 8.3.13 (api:88/proto:86-96)
GIT-hash: 83ca112086600faacab2f157bc5a9324f7bd7f77 build by mockbuild@builder10.centos.org, 2012-05-07 11:56:36
1: cs:Connected ro:Secondary/Primary ds:UpToDate/UpToDate C r-----
ns:41 nr:2377 dw:2379 dr:173 al:1 bm:1 lo:0 pe:0 ua:0 ap:0 ep:1 wo:b oos:0
结论：说明主备上DRBD启动正常，之前连接正常！
2.启动heartbeat，通过日志查看heartbeat包括关联的服务加载的情况，tail -f /var/log/ha.log
[root@mfs-master /]# tail -f /var/log/ha.log
IPaddr[15336]:2012/06/05_19:26:04 INFO: Resource is stopped
ResourceManager[15238]:2012/06/05_19:26:04 info: Running /etc/ha.d/resource.d/IPaddr 192.168.8.210 start
IPaddr[15409]:2012/06/05_19:26:04 INFO: Using calculated nic for 192.168.8.210: eth0
IPaddr[15409]:2012/06/05_19:26:04 INFO: Using calculated netmask for 192.168.8.210: 255.255.255.0
IPaddr[15409]:2012/06/05_19:26:04 INFO: eval ifconfig eth0:0 192.168.8.210 netmask 255.255.255.0 broadcast 192.168.8.255
IPaddr[15392]:2012/06/05_19:26:04 INFO: Success
heartbeat[15225]: 2012/06/05_19:26:04 info: local HA resource acquisition completed (standby).
heartbeat[15197]: 2012/06/05_19:26:04 info: Standby resource acquisition done [foreign].
heartbeat[15197]: 2012/06/05_19:26:04 info: Initial resource acquisition complete (auto_failback)
heartbeat[15197]: 2012/06/05_19:26:05 info: remote resource transition completed.
结论：heartbeat 启动成功！
3.在主机器上ps的方式查看heartbeat 和MFS进程，确保服务启动，如果没有启动，可手动在启动一次。此时备份机器的mfs没有启动。
[root@mfs-master /]# ps aux |grep haert
root 15507 0.0 0.2 61220 748 pts/0 S+ 19:30 0:00 grep haert
[root@mfs-master /]# ps aux |grep heart
root 15197 0.0 4.4 58008 13008 ? SLs 19:25 0:00 heartbeat: master control process
nobody 15200 0.0 2.1 51412 6412 ? SL 19:25 0:00 heartbeat: FIFO reader
nobody 15201 0.0 2.1 51408 6408 ? SL 19:25 0:00 heartbeat: write: bcast eth1
nobody 15202 0.0 2.1 51408 6408 ? SL 19:25 0:00 heartbeat: read: bcast eth1
root 15509 0.0 0.2 61220 744 pts/0 S+ 19:30 0:00 grep heart
[root@mfs-master /]# ps aux |grep mfs
root 14914 0.0 3.5 136048 10416 ? S 19:18 0:00 python /data/mfs/sbin/mfscgiserv
mfs 15320 0.8 27.3 93880 80508 ? S< 19:25 0:02 /usr/local/mfs/sbin/mfsmaster start
root 15511 0.0 0.2 61220 744 pts/0 S+ 19:30 0:00 grep mfs
结论：主服务器上各服务正常启动
4.停止主服务的heartbeat，查看主服务的日志，发现所有的关联服务都会被关闭，包括drbd,mfs,vip等。
[root@mfs-master mfs]# service heartbeat stop
Stopping High-Availability services:
[ OK ]
[root@mfs-master mfs]# tail -f /var/log/ha.log
ResourceManager[14246]:2012/06/05_19:06:57 info: Running /etc/ha.d/resource.d/mfsmaster stop
ResourceManager[14246]:2012/06/05_19:06:57 info: Running /etc/ha.d/resource.d/Initdrbd stop
heartbeat[14233]: 2012/06/05_19:06:58 info: All HA resources relinquished.
heartbeat[8527]: 2012/06/05_19:07:00 info: killing HBFIFO process 8529 with signal 15
heartbeat[8527]: 2012/06/05_19:07:00 info: killing HBWRITE process 8530 with signal 15
heartbeat[8527]: 2012/06/05_19:07:00 info: killing HBREAD process 8531 with signal 15
heartbeat[8527]: 2012/06/05_19:07:00 info: Core process 8530 exited. 3 remaining
heartbeat[8527]: 2012/06/05_19:07:00 info: Core process 8531 exited. 2 remaining
heartbeat[8527]: 2012/06/05_19:07:00 info: Core process 8529 exited. 1 remaining
heartbeat[8527]: 2012/06/05_19:07:00 info: mfs-master Heartbeat shutdown complete.
结论：可以看到各关联服务正关闭！