对基于jboss的服务辅助监控

对基于jboss服务辅助监控

 

脚本部署在10.247crontab中中的部署：

# Job log update monitor

*/10   *   *  *  *   /usr/local/shell/web-tools/service-monitor/update-monitor.sh > /dev/null 2>&1

*/20   *   *  *   *  /usr/local/shell/web-tools/service-monitor/spiderjob.sh > /dev/null 2>&1

*/40   *   *  *   *  /usr/local/shell/web-tools/service-monitor/expjob.sh > /dev/null 2>&1

 

监控对象：

（1）10.190 10.80 10.82 10.83 10.85 10.87 30.11 30.12 30.101

（2）40.31 40.32

（3）40.72

 

监控原理：

           相关脚本每10分钟或20分钟或40分钟检测监控对象的jboss日志大小是否增加，来辅助判断基于jboss的服务是否正常。这种方法不能一锤子定音，但可以作为判断服务是否正常的辅助手段。

报警：

  如果检测日志大小不大于上次检测日志的大小，就会连接10.247报警平台短信报警,报警内容(1)、ip地址 、 Job log have NOT updated in 10 min、报警时间.(2)、ip地址、SpiderJob log have NOT updated in 20 min、报警时间(3)、ip地址、Job log have NOT updated in 40 min、报警时间。

 

脚本源码：

(1)/usr/local/shell/web-tools/service-monitor/update-monitor.sh

#!/bin/sh

# monitor the logs' size

# Shao

 

time=`date +%T`

hour=`date +%H`

min=`date +%M`

flushtime=$hour$min

 

echo $flushtime

 

SMhome="/usr/local/shell/web-tools/service-monitor"

tmpdir="$SMhome"/tmp

 

jbosslog=/usr/local/jboss/server/default/log/server.log

nginxlog=nginxlog=/usr/local/nginx/logs/

apachelog=/usr/local/httpd/logs/access_log

 

 

 

 

#for ip in 10.190 10.80 10.82 10.83 10.85 10.87 30.11 30.12 30.101 40.31 40.32 40.33 ; do

for ip in 10.190 10.80 10.82 10.83 10.85 10.87 30.11 30.12 30.101; do

    echo $ip

 

    jlogtmp=$tmpdir/$ip.jlog.tmp

 

    #flush

    if ; then

         echo 00 > $tmpdir/$ip.tmp

    fi

 

    size=`ssh 10.10.$ip "ls -l $jbosslog" | awk '{print $5}'`

    lastsize=`tail -2 $jlogtmp|head -1`

    echo $size >> $jlogtmp

 

# log update

    echo $size $lastsize

    if [ $size -eq $lastsize ]; then

        ssh 10.10.10.247 "fetion.sh phone="13811371488,13811193602" msg="$ip: Job log have NOT updated in 10 min.$time."" > /dev/null

        #wget  --output-document=/dev/null " Job log have NOT updated in 20 min.$time."

    fi

 

done

#####################################################################

（2）、#!/bin/sh

# monitor the logs' size

# Shao

 

time=`date +%T`

hour=`date +%H`

min=`date +%M`

flushtime=$hour$min

 

echo $flushtime

 

SMhome="/usr/local/shell/web-tools/service-monitor"

tmpdir="$SMhome"/tmp

 

jbosslog=/usr/local/jboss/server/default/log/server.log

nginxlog=nginxlog=/usr/local/nginx/logs/

apachelog=/usr/local/httpd/logs/access_log

 

 

 

 

for ip in 40.31 40.32; do

    echo $ip

 

    jlogtmp=$tmpdir/$ip.jlog.tmp

 

    #flush

    if ; then

         echo 00 > $tmpdir/$ip.tmp

    fi

 

    size=`ssh 10.10.$ip "ls -l $jbosslog" | awk '{print $5}'`

    lastsize=`tail -2 $jlogtmp|head -1`

    echo $size >> $jlogtmp

 

# log update

    echo $size $lastsize

    if [ $size -eq $lastsize ]; then

        ssh 10.10.10.247 "fetion.sh phone="13811371488,13811193602" msg="$ip: SpiderJob log have NOT updated in 20 min.$time."" > /dev/null

        #wget  --output-document=/dev/null " Job log have NOT updated in 20 min.$time."

    fi

 

done

 

#########################################################################

 

（3）、/usr/local/shell/web-tools/service-monitor/expjob.sh

#!/bin/sh

# monitor the logs' size

# Shao

 

time=`date +%T`

hour=`date +%H`

min=`date +%M`

flushtime=$hour$min

 

echo $flushtime

 

SMhome="/usr/local/shell/web-tools/service-monitor"

tmpdir="$SMhome"/tmp

 

jbosslog=/usr/local/jboss/server/default/log/server.log

nginxlog=nginxlog=/usr/local/nginx/logs/

apachelog=/usr/local/httpd/logs/access_log

 

for ip in 40.72; do

#for ip in 10.190 10.80 10.82 10.83 10.85 10.87 30.11 30.12 30.101 40.72; do

    echo $ip

 

    jlogtmp=$tmpdir/$ip.jlog.tmp

 

    #flush

    if ; then

         echo 00 > $tmpdir/$ip.tmp

    fi

 

    size=`ssh 10.10.$ip "ls -l $jbosslog" | awk '{print $5}'`

    lastsize=`tail -2 $jlogtmp|head -1`

    echo $size >> $jlogtmp

 

# log update

    echo $size $lastsize

    if [ $size -eq $lastsize ]; then

        ssh 10.10.10.247 "fetion.sh phone="13811371488,13811193602,15010801866" msg="$ip: Job log have NOT updated in 40 min.$time."" > /dev/null

        #wget  --output-document=/dev/null " Job log have NOT updated in 20 min.$time."

    fi

 

done