Apache网站集群corosync+pacemaker架构使用crm_mon做简单监控 推荐

思路：使用pacemaker原生的CLI管理工具组件的crm_mon，监控集群的状态

梗概：把crm_mon做成LSB标准的服务。使用crm configure工具添加到集群资源中。设置Apache资源与crm_mon资源的启动顺序和绑定关系。

操作：

一、制作标准的LSB 架构crm_mon服务以方便pacemaker管理，拷贝如下脚本到集群节点的/etc/init.d/目录下并给执行权限

注：可以根据实际情况，修改crm_mon的参数，例如我们还可以使用

crm_mon --daemonize --snmp-traps snmptrapd.example.com

来发出SNMP信息，并通过nagios等SNMP代理收集并处理。

#!/bin/bash
#
# crm_mon        Startup script for the Pacemaker managment tools: crm_mon
#
# chkconfig: - 86 16
# description: crm_mon scripts for pacemaker's lsb style
# processname: crm_mon
# pidfile: /var/run/crm_mon.pid
#
### BEGIN INIT INFO
### END INIT INFO
# wangxiaoyu#live.com
# 2011-03-20 22:53:59
# Version 0.1

# Source function library.
. /etc/rc.d/init.d/functions

crm_mon=/usr/sbin/crm_mon
prog=crm_mon
pidfile=${PIDFILE-/var/run/crm_mon.pid}
lockfile=${LOCKFILE-/var/lock/subsys/crm_mon}
RETVAL=0

# 请根据实际情况，修改crm_mon的启动参数
OPTIONS='--daemonize --as-html /var/www/html/crm_mon.html'

start() {
echo -ne $"Starting $prog: "
$crm_mon $OPTIONS
RETVAL=$?
[ $RETVAL -ne 0 ] && echo_failure && return $RETVAL
pidNum=$(pidof ${crm_mon})
[ "$pidNum" -ne 1 ] && touch ${pidfile} \
&& echo -ne ${pidNum} > ${pidfile} \
&& touch ${lockfile}
RETVAL=$?
[ $RETVAL = 0 ] && echo_success
return $RETVAL
# 最初尝试通过function中的deamon函数实现，发现不成功，只好自己写pidfile和lockfile
}

# When stopping crm_mon a delay of >10 second is required before SIGKILLing the
# crm_mon parent; this gives enough time for the crm_mon parent to SIGKILL any
# errant children.
stop() {
echo -n $"Stopping $prog: "
killproc -p ${pidfile} -d 10 $crm_mon
RETVAL=$?
echo
[ $RETVAL = 0 ] && rm -f ${lockfile} ${pidfile}
}
reload() {
echo -n $"Reloading $prog: "
# Force LSB behaviour from killproc
LSB=1 killproc -p ${pidfile} $crm_mon -HUP
RETVAL=$?
if [ $RETVAL -eq 7 ]; then
failure "$crm_mon shutdown"
fi
echo
}

# See how we were called.
case "$1" in
start)
start
;;
stop)
stop
;;
status)
status -p ${pidfile} $crm_mon
RETVAL=$?
;;
restart)
stop
start
;;
condrestart|try-restart)
if status -p ${pidfile} $crm_mon >&/dev/null; then
stop
start
fi
;;
force-reload|reload)
reload
;;
*)
echo $"Usage: $prog {start|stop|restart|condrestart|try-restart|force-reload|reload|status|help}"
RETVAL=2
esac

exit $RETVA

二、添加crm_mon资源到集群中，并设置与资源HTTPD（管理Apache服务的脚本）的绑定关系和启动顺序

在任意一集群节点上操作：

crm configure primitive CRM_MON lsb:crm_mon

crm configure colocation CRM_MON_with_HTTPD inf: CRM_MON HTTPD

crm configure order order HTTPD_before_CRM_MON inf: HTTPD CRM_MON

三、现在我们就可以通过http://VIP/crm_mon.html页面实时查看集群的状态了

# elinks http://VIP/crm_mon.html # 结果如下

Cluster summary

Last updated: Sun Mar 20 22:32:38 2011

Current DC: pcmk-2 (pcmk-2)

3 Nodes configured.

3 Resources configured.

Config Options

STONITH of failed nodes	:	disabled
Cluster is	:	symmetric
No Quorum Policy	:	Ignore

Node List

Node: pcmk-2 (pcmk-2): online

Node: pcmk-1 (pcmk-1): standby

Node: pcmk-3 (pcmk-3): online

Resource List

VIP (ocf::heartbeat:IPaddr2): Started pcmk-2

HTTPD (ocf::heartbeat:apache): Started pcmk-2

CRM_MON (lsb:crm_mon): Started pcmk-2

附件：http://down.51cto.com/data/2357781