IBM Mainframe Forum

Is it possible to join 2 files based on multiple fields ?
So far I have used only one field to do the join on 2 files .

Also the how would be a performance if we are comparing the 2 files having a million records each .

I would appreciate if simple example is posted to demonstrate how the join based on multiple fields work .

Joining on multiple fields works the same as joining on a single field.
For example:
If you need further assistance, then please provide more detailed information regarding your requirements, such as RECFM and LRECL of each input file, position and lengths of KEY fields to be JOINed, and what you want included in the output records.

Thanks for your reply .

I have around 8-10 fields which can be used in Join .

I was curious about performance .
How it would work when 2 files involved have million rows each .

Its not the number of records in each file that you need to focus on, but rather the number of duplicate keyed records. The file with more duplicate keyed records should be coded as SORTJNF1.