vzmz
You can strip the header records of with a combination head and tail commands or sed. Do a search. wc -l filename will tell you how many lines are in the file. These commands are very fast. So in a shell script I would validate the file and start the DataStage job after it passes maybe move the bad files to a separate folder. You should archive the good files when done to a different folder.
Kim.
Set Of files
Moderators: chulett, rschirm, roy