IBM Mainframe Forum

Old file is unload of production database
Old file was used to load test database
New file is unload of that test database before any changes have been made to it

FILE-AID COMPARE is showing that records have been changed

Tried using COMPARE TYPE = S and using different locations in the record as the sort key. Every instance me an error message

U326-RECORD IS TOO SHORT TO CONTAIN KEY

which I have not been able to locate in manuals or online. And it read only 3 records when I told it to read and compare 25 records.
OK, I may have figured that part out. Because it was VBS record I was subtracting the 4 bytes from the column location. Instead I used the
actual column location that was showing and that helped.

But now the problem seems to be that because there are records of varying lengths on the file I am not able to sort the shorter records from the
longer records. And so the two files are not in the exact same sequence. The longer record has a description and that is what is getting
out of sequence.

How can I get the records sorted into the correct order? I tried finding a sequence in the "key" information at the start of the record but haven't had any luck with that either.

We have a very similar need in my shop. We ended up writing a "preprocessor" that prefixed each record with its key (which was not necessarily composed in the same way for every record in a given unload) and sorted on that.