원문 : http://www.ischo.net -- 조인상 //시스템 엔지니어

+++++++++++++++++++++++++++++++++++++++++++++++++++++++

프로세스 생사여부 검사후 메일보내는 쉘스크립트

먼저 해당 디렉토리의  proclist.txt 파일에 정의된 프로세스 리스트들을 읽어서

생사여부를 검사하고

죽은 프로세스가 있으면 메일로 보내주는 스크립트

완전하진 않지만 그럭저럭 쓸만함.

#!/bin/ksh

######################################################################################
# script by ischo.
# http://www.ischo.net
# mail:chosim1@naver.com
# script description :
# This script read first $PROG_HOME/proclist.txt file, then check process.
# If exist died process, make Mail Contents and send mail to $MAIL_RECIEVER
######################################################################################

######################################################################################
## ENVIRONMENT SETTING - You can edit this.
######################################################################################

# Absolute Path of script
PROG_HOME=/root/chkproc

# E-mail reciever
MAIL_RECIEVER=chosim1@naver.com

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

######################################################################################
# Variables for host environment - You do not need to edit.
######################################################################################

# ServerName : This setting will be included in SMS
SERVERNAME=`hostname`

# Config File List
MAILFILE=${PROG_HOME}/chkproc.mail
LOGFILE=${PROG_HOME}/chkproc_`date +%Y%m%d`.log
PROCLIST=${PROG_HOME}/proclist.txt
ALIVEFILE=${PROG_HOME}/alive_proclist.tmp
LASTSTATFILE=${PROG_HOME}/laststat.tmp
CURSTATFILE=${PROG_HOME}/curstat.tmp
TEMPSTATFILE=${PROG_HOME}/tempstat.tmp

# TIME mark VARS
STIME=`date +%Y%m%d%H%M%S`
TTIME=`date +%H:%M`
DDATE=`date +%Y%m%d`

# First VAR's value for Operation
ERROR=0
NOPROC=`cat ${PROCLIST} | wc -l`
Recover_stat="no"
Down_stat="no"

# First creation of LASTSTATFILE
if [ -f ${LASTSTATFILE} ];
then
        touch ${LASTSTATFILE}
else
        for TEMP in `cat ${PROCLIST}`
        do
        echo "${TEMP} : DOWN" >> ${LASTSTATFILE}
        done
fi

cat /dev/null > ${CURSTATFILE}

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

######################################################################################
# DEFINE FUNCTIONS
######################################################################################
# Function SendMail - Sending Mail of Error
function Func_SendMail
{

for PROCNAME in `cat ${PROCLIST}`
do

        CURRENT_STAT=`cat ${CURSTATFILE} | grep ${PROCNAME} | awk '{print $3}'`
        LAST_STAT=`cat ${LASTSTATFILE} | grep ${PROCNAME} | awk '{print $3}'`

        if [ ${CURRENT_STAT} = UP ];
        then
                if [ ${LAST_STAT} = DOWN ];
                then
                Recover_stat="yes"
                fi
                cat ${LASTSTATFILE} | grep -v ${PROCNAME} > ${TEMPSTATFILE}
                cat ${TEMPSTATFILE} > ${LASTSTATFILE}
                echo "${PROCNAME} : UP" >> ${LASTSTATFILE}
        else
                if [ ${LAST_STAT} = UP ];
                then
                Down_stat="yes"
                fi
                cat ${LASTSTATFILE} | grep -v ${PROCNAME} > ${TEMPSTATFILE}
                cat ${TEMPSTATFILE} > ${LASTSTATFILE}
                echo "${PROCNAME} : DOWN" >> ${LASTSTATFILE}
        fi

done

if [ ${Down_stat} = yes ];
then
mail -s "Process Down Alert!" ${MAIL_RECIEVER} < ${MAILFILE}
fi

if [ ${Recover_stat} = yes ];
then
echo "mail -s 'Process Recovered!' ${MAIL_RECIEVER} < ${MAILFILE}"
mail -s "Process Recovered!" ${MAIL_RECIEVER} < ${MAILFILE}
fi

}

# Function LogTop - Logging Number of Alive Processes
function Func_LogTop
{
NOP=0
cat /dev/null > ${ALIVEFILE}

for TEMP_PROC in `cat ${PROCLIST}`
do

        for REAL_PROC in `ps -ef | grep ${TEMP_PROC} | grep -v grep | awk {'print $8'} | sort | uniq`
        do
                if [ ${TEMP_PROC} = ${REAL_PROC} ];
                then
                ps -ef | grep ${TEMP_PROC} | grep -v grep | awk {'print $8'} | sort | uniq >> ${ALIVEFILE}
                NOP=`expr ${NOP} + 1`
                fi
        done

done

echo "${DDATE} ${TTIME}  Number of alive Processes : ${NOP}/${NOPROC}" >> ${LOGFILE}
}

# Function Logging - Log Process Check Result
function Func_Logging
{
echo "${DDATE} ${TTIME}  --Process ${PROC} Down" >> ${LOGFILE}
}

# Function LogBottom - Log Bottom Message
function Func_LogBottom
{
cat /dev/null > ${ALIVEFILE}

if [ ${ERROR} -eq 1 ];
then
echo "ps -ef result :" >> ${LOGFILE}
cat ${ALIVEFILE}  >> ${LOGFILE}
fi

}

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

#############################################################################################
# Main Procedure
# - Start Program
#############################################################################################

echo "-- ${DDATE} ${TTIME}  ${SERVERNAME} Process Check List --" > $MAILFILE

# Logging Number of Alive Processes
Func_LogTop

# Loop Process check and Log
for PROC in `cat $PROCLIST`
do

PROC_CHECK=`ps -ef | grep $PROC | grep -v grep | wc -l`
SEND_MSG=`echo "[critical] $SERVERNAME Process $PROC Down. $TTIME"`

        if [ $PROC_CHECK -eq 0 ];
        then

        echo "-- Process $PROC Down. ${TTIME}" >> $MAILFILE
        echo "${PROC} : DOWN" >> ${CURSTATFILE}
        Func_Logging
        ERROR=1

        else
        {
        echo "-- Process $PROC alive. ${TTIME}" >> $MAILFILE
        echo "${PROC} : UP" >> ${CURSTATFILE}
        }

        fi

done

# Send Mail
Func_SendMail

# Maintain Number of LOG files
find ${PROG_HOME} -name "*.log" -type f -mtime +5 -exec rm {} \;