awk command with multiple commands

There is usualy no need to grep .. | awk ..| awk

I change grep "Mem" | awk '{print 0 + $7}' | awk '{ print $1 / 1024 }'

• to awk '/Mem/ {print 0 + $7}' | awk '{ print $1 / 1024 }'
• to awk '/Mem/ {print 0 + $7/1024 }'
• to awk '/Mem/ { if ( $7 > 1024 ) ...

I would start with

top -M -n1 | awk '/Mem/ {if ($7 > 1024) { 
      print "Memory utilisation is high \n" ;
      print "Please find history of the memory utilisation below\n" ;
      print " sar -r|awk \'{print $1,$2,$3,$4}\'| column -t \" } }' >>/home/shyam/utilisation.txt

What does top -M -n 1 do? On my system, -M is an invalid option. I assume you're trying to get the amount of memory in use on the system (in megabytes)? In that case, use free rather than top - extracting data from an ncurses app like top is silly when there are plain-text tools like free to do the job.

Anyway, I'm going to use free -m in my example script.

Building on Archemar's work to eliminate excess grep and awk commands:

mem=$(free -m | awk '/Mem:/ {print $3}')

if [ "$mem" -gt 1024 ] ; then (
    echo "Memory utilisation is high"
    echo "Please find history of the memory utilisation below"
    sar -r | awk '{print $1,$2,$3,$4}' | column -t
) >>/home/shyam/utilisation.txt
fi

Instead of using awk to do all of the work, it only uses awk to extract the memory usage data from the output of free -m, and then it uses shell code to do the rest. It runs echo and sar | awk inside a sub-shell, with the entire output of the sub-shell redirected to utilisation.txt

If you don't have free on your system and top -M -n 1 works for you then use this as the first line:

mem=$(top -M -n1 | awk '/Mem/ {print $7}')