(copy) Shell Script to Check Linux System Health

source: http://linoxide.com/linux-shell-script/shell-script-check-linux-system-health/

This article introduces a shell script to perform linux system health check.

This script collects system information and status like hostname, kernel version, uptime, cpu / memory / disk usage.
Script uses:
hostname, uptime, who, mpstat, lscpu, ps, top, df, free, bc commands to get system information
and cut, grep, awk and sed for text processing.

The output of the script is a text file which will be generated in the current directory.

A variable is set to provide email address to which script can send report file.

Apart from system status, the script will check a predefined threshold for cpu load and filesystem size.

Remember : Make sure you have all the above commands working, to output all results correctly.

 #!/bin/bash
 EMAIL=''
 function sysstat {
 echo -e "
 #####################################################################
     Health Check Report (CPU,Process,Disk Usage, Memory)
 #####################################################################

 Hostname         : `hostname`
 Kernel Version   : `uname -r`
 Uptime           : `uptime | sed 's/.*up \([^,]*\), .*/\1/'`
 Last Reboot Time : `who -b | awk '{print $3,$4}'`

 *********************************************************************
 CPU Load - > Threshold <  Normal >  Caution , >  Unhealthy
 *********************************************************************
 "
 MPSTAT=`which mpstat`
 MPSTAT=$?
 if [ $MPSTAT !=  ]
 then
     echo "Please install mpstat!"
     echo "On Debian based systems:"
     echo "sudo apt-get install sysstat"
     echo "On RHEL based systems:"
     echo "yum install sysstat"
 else
 echo -e ""
 LSCPU=`which lscpu`
 LSCPU=$?
 if [ $LSCPU !=  ]
 then
     RESULT=$RESULT" lscpu required to producre acqurate reults"
 else
 cpus=`lscpu | grep -e "^CPU(s):" | cut -f2 -d: | awk '{print $1}'`
 i=
 while [ $i -lt $cpus ]
 do
     echo "CPU$i : `mpstat -P ALL | awk -v var=$i '{ if ($3 == var ) print $4 }' `"
     let i=$i+
 done
 fi
 echo -e "
 Load Average   : `uptime | awk -F'load average:' '{ print $2 }' | cut -f1 -d,`

 Heath Status : `uptime | awk -F'load average:' '{ print $2 }' | cut -f1 -d, | awk '{if ($1 > 2) print "Unhealthy"; else if ($1 > 1) print "Caution"; else print "Normal"}'`
 "
 fi
 echo -e "
 *********************************************************************
                              Process
 *********************************************************************

 => Top memory using processs/application

 PID %MEM RSS COMMAND
 `ps aux | awk '{print $2, $4, $6, $11}' | sort -k3rn | head -n `

 => Top CPU using process/application
 `top b -n1 | head - | tail -`

 *********************************************************************
 Disk Usage - > Threshold <  Normal > % Caution >  Unhealthy
 *********************************************************************
 "
 df -Pkh | grep -v 'Filesystem' > /tmp/df.status
 while read DISK
 do
     LINE=`echo $DISK | awk '{print $1,"\t",$6,"\t",$5," used","\t",$4," free space"}'`
     echo -e $LINE
     echo
 done < /tmp/df.status
 echo -e "

 Heath Status"
 echo
 while read DISK
 do
     USAGE=`echo $DISK | awk '{print $5}' | cut -f1 -d%`
     if [ $USAGE -ge  ]
     then
         STATUS='Unhealty'
     elif [ $USAGE -ge  ]
     then
         STATUS='Caution'
     else
         STATUS='Normal'
     fi

         LINE=`echo $DISK | awk '{print $1,"\t",$6}'`
         echo -ne $LINE "\t\t" $STATUS
         echo
 done < /tmp/df.status
 rm /tmp/df.status
 TOTALMEM=`free -m | head - | tail -| awk '{print $2}'`
 TOTALBC=`echo "scale=2;if($TOTALMEM<1024 && $TOTALMEM > 0) print 0;$TOTALMEM/1024"| bc -l`
 USEDMEM=`free -m | head - | tail -| awk '{print $3}'`
 USEDBC=`echo "scale=2;if($USEDMEM<1024 && $USEDMEM > 0) print 0;$USEDMEM/1024"|bc -l`
 FREEMEM=`free -m | head - | tail -| awk '{print $4}'`
 FREEBC=`echo "scale=2;if($FREEMEM<1024 && $FREEMEM > 0) print 0;$FREEMEM/1024"|bc -l`
 TOTALSWAP=`free -m | tail -| awk '{print $2}'`
 TOTALSBC=`echo "scale=2;if($TOTALSWAP<1024 && $TOTALSWAP > 0) print 0;$TOTALSWAP/1024"| bc -l`
 USEDSWAP=`free -m | tail -| awk '{print $3}'`
 USEDSBC=`echo "scale=2;if($USEDSWAP<1024 && $USEDSWAP > 0) print 0;$USEDSWAP/1024"|bc -l`
 FREESWAP=`free -m |  tail -| awk '{print $4}'`
 FREESBC=`echo "scale=2;if($FREESWAP<1024 && $FREESWAP > 0) print 0;$FREESWAP/1024"|bc -l`

 echo -e "
 *********************************************************************
              Memory
 *********************************************************************

 => Physical Memory

 Total\tUsed\tFree\t%Free

 ${TOTALBC}GB\t${USEDBC}GB \t${FREEBC}GB\t$(($FREEMEM *  / $TOTALMEM  ))%

 => Swap Memory

 Total\tUsed\tFree\t%Free

 ${TOTALSBC}GB\t${USEDSBC}GB\t${FREESBC}GB\t$(($FREESWAP *  / $TOTALSWAP  ))%
 "
 }
 FILENAME="health-`hostname`-`date +%y%m%d`-`date +%H%M`.txt"
 sysstat > $FILENAME
 echo -e "Reported file $FILENAME generated in current directory." $RESULT
 if [ "$EMAIL" != '' ]
 then
     STATUS=`which mail`
     if [ "$?" !=  ]
     then
         echo "The program 'mail' is currently not installed."
     else
         cat $FILENAME | mail -s "$FILENAME" $EMAIL
     fi
 fi