[saccskiz]

I am on a Sun Solaris 5.8
A piece of my load.ksh script has the following lines :

cat /data/dev/CODE_LIST1.TXT | while read column1 column2
do
genscript.ksh ${dbname} ${schemaname} member $column1 /data/dev/member_scripts/$column2
done

The CODE_LIST1.TXT looks something like

38222 38222_123456789.sql
48338 48338_438434845.sql


I am reading a CODE_LIST1.TXT file and passing the contents as variables into the genscript.ksh script
The problem is that CODE_LIST1.TXT could have 100 lines out there and currently our process if there are
100 entries in the .TXT file takes about 6 unavoidable hours just due to the processing involved.

Can I get some pointers on how I can multi-thread this code so that if there are more than 4 entries
in the CODE_LIST1.TXT, I want to split it up into 4 execution threads - each picking one up 1 line.
If there are 100 lines, each of the parallel threads need to pick up 25 each. If there are 105 lines,
then, first 3 picks up 27 each and the last one picks up 24 lines to process.
Can this be done ?
Thanks

[saccskiz]

I figured this out...

linecount=`cat CODE_LIST1.TXT |wc -l`
echo $linecount
splitcount=`echo "scale=0; $linecount/4+1" |bc -l`
echo $splitcount
split -l $splitcount CODE_LIST1.TXT

## By the above, I now have 4 files - xaa,xab,xac and xad

for f in x*
do
parallelproc.ksh $dbname $schemaname $f &
done
wait


And in my parallelproc.ksh I have :

#!/bin/ksh
dbname=$1
schemaname=$2
filename=$3
cat /data/dev/$filename | while read column1 column2
do
genscript.ksh ....(with other parameters as in original posted thread)
done
exit