IBM Mainframe Forum

If you have a one-to-one match on both files and they are both in the same order then you do not need a key.

k
I'll try..

Do you have duplicates on the keys, I asked. Either file.

I was checking that. First 30 and last 30 IDs of both files are exactly matching. One file has 2 extra header lines. Still both files have same record count. I don't know why..

OK, first thing, simple JOINKEYS on the keys. You have a "test" extract I assume, which contains representative data?

I'm trying, Taking time because of x-wait...

dfsort-icetool-icegener/topic6864.html

Isn't this post similar to ours?

NO
this is ( in case You forgot ) about file comparison
that one is about improving performance for a large sort

Ok

Since all IDs in both files are unique and same, I'm trying to do one-to-one matching. And will run compare twice.
Before I was able to compare first 3000 bytes of 7000 bytes by using this:

//DFSPARM DD *
OPTION DYNALLOC=(SYSDA,70),AVGRLEN=7000
/*
//TOOLIN DD *
SELECT FROM(SORTIN) TO(OUTPUT) ON(1,3000,CH) NODUPS
/*


Now I thought I will compare for first 3500 bytes like this:
//DFSPARM DD *
OPTION DYNALLOC=(SYSDA,200)
/*
//TOOLIN DD *
SELECT FROM(SORTIN) TO(OUTPUT) ON(1,3500,CH) NODUPS
/*

Eventhough allocating 200 SYSDA's it shows:
ICE046A 0 SORT CAPACITY EXCEEDED