I have a requirement like this. Thru one of the program a huge file is splitted into 'N' number of versions with a unique sequence number to each version.
In the next step after creating versions, i have to take the first version and process it.
Here the versions created are not constant, if so we can direct our jcl to take (-n) of that gdg. For us the sequence number is very much essential if first version has seq no as 1 and 10th version wil have seq no of 10 and i need to process this in an order.
Joined: 14 Mar 2007 Posts: 8686 Location: Back in jolly old England
Why not use fixed dataset names rather than a gdg
The terms versions and generations are both valid in relationship to a GDS. As only one version of a GDS can exist at any given time, I can only assume that you really meant to say generations.
Please learn to use the correct terminology when describing your problem. Take the time to learn what is right and make things easier and clearer for those who try to help you