MVSFORUMS.com

kirank

Hi I've 2 files and first file is having around 10 million records unique records. The length of the keys starts from 1 to 23 bytes. And the second file contains 340 million records and first 23 bytes contains key values but they contain duplicates. I want to read the record from the first file and match the records from the second file and write all the matched records from the second file.

I've used syncsort using joinkeys but the job is abending due to buffers unavailable. I've increased sort work areas and used SDB=LARGE. Still I'm getting abend. Please suggest an alternate approach. Thank you

William Collins · Supermod Joined: 03 Jun 2012 Posts: 437 Topics: 0

Why are you using SORTWK datasets? You don't have a 340-million-record file with a 23-byte key and it is not in order?

If the data is in order, specify SORTED on the JOINKEYS statement for that file.

If for some unusual reason your data is not in order, post the sysout from the failing step.

kirank · Posted: Sat Apr 25, 2015 10:52 pm Post subject:

Both the files are in same sorted order. I will post the abend details. Thank you

William Collins · Supermod Joined: 03 Jun 2012 Posts: 437 Topics: 0

kirank · Posted: Tue Apr 28, 2015 10:55 am Post subject:

Hi the problem is solved. I've increased VOL=(,,,6) to VOL=(,,,26) and mentioned SORTED for second file and the job executed successfully. Thank you for all your help Smile

William Collins · Supermod Joined: 03 Jun 2012 Posts: 437 Topics: 0

Since you said both files were in sequence already, you're still wasting a lot by not specifying SORTED for both the JOINKEYS files.