num variable is not getting identified in the below script correctly and so it throws me this error ( sed: command garbled: $num p
No message !?! )

#!/bin/ksh
ps -e | grep 'oracle' | nawk '{print $1, $3}' > /check/oracle.log
sed -e 's/\:[0-9][0-9]//g' /check/oracle.log | nawk 'BEGIN { OFS = "\t" }; {print $1, $2}' > /check/oracle1.log
wc -l /check/oracle1.log | nawk '{print $1}' > /check/count.log
COUNT=`cat /check/count.log`
$1}' > /dev/null`
CHK=`cat /check/oracle2.log | nawk '{print $2}'`
num=1
while (( num <= $COUNT ))
do
sed -n -e '$num p' /check/oracle1.log > /check/oracle2.log
if [ $CHK >= 300 ]
then
cat /check/oracle2.log | nawk '{print $1}' | mailx user@domain.com
#else
#exit 1
fi
(( num = num + 1 ))
done
print "Done."

Can anyone help me out with this, please?

Aladimi

replace: (single quotes)

sed -n -e '$num p' /check/oracle1.log > /check/oracle2.log

with: (double quotes)

sed -n -e "$num p" /check/oracle1.log > /check/oracle2.log

Luke Chi