LISTCAT LEVEL(HLQ) VOLUME cc=8 not sufficient region 8M

Douglas Wilder · Posted: Fri Jan 19, 2007 11:07 pm

We have a batch job that uses ID-CAMS "LISTCAT LEVEL(HLQ) VOLUME" then a utility to separate out the datasets on volumes with a given prefix. The LISTCAT has started ending with a condition code 8 and a message that it does not have sufficient region. We have set the region up to 8M and still occasionally have the problem. The region size needed seems to be related to the number of datasets retrieved by the LISTCAT. Is there some other way to get all of the dataset names on volumes with a given prefix that would not require ever expanding region.

dick scherrer · Posted: Sat Jan 20, 2007 12:19 am

Hello,

Depending on how how your hlq is structured, you could split the listcat steps into multiples and mod onto the output. Your "utility" would run from basically the same input it has now - with maybe a few extra header pages.

I believe IDCAMS will "grow" as the number of datasets increases.

If your site allows, you might also try REGION=0M (zeroM).

Douglas Wilder · Posted: Tue Jul 03, 2007 11:47 pm

I would like to belatedly thank you for your prior help.
I did not fine a way to break it up into a reasonable number of groups, so I coded region=0M like you suggested. That worked fine, until now.

We are again getting "IDC3332I ** INSUFFICIENT MAIN STORAGE" even with region=0M.

Any other good ideas? This part of the job is trying to get a list of the DSN and VOL SER of all of our cataloged carts and tapes. Is there some other way to get this list without infinite region. Do you have any Idea why it requires so much region?

dick scherrer · Posted: Wed Jul 04, 2007 12:17 am

You're welcome

What is the HLQ you are "seeding" the listcat with? If, for example, you have started at the highest level, you may need to create multiple listcats of the first 2 levels.

As more and more entries are made into your catalog, you will always be able to create more names than can be processed in a single run.

Douglas Wilder · Posted: Wed Jul 04, 2007 4:12 am

This job is using a HLQ of just the highest node like ABC. It is my understanding that a complete node is required in the LEVEL. I found that there are over 1000 different second nodes that start with a particular letter ie ABC.D*.** there would be many more second nodes starting with other letters. How I could break this up into a reasonable number of listcats? Are there any other utilities that I might be able to use instead of listcat that might not use so much region?

expat · Posted: Wed Jul 04, 2007 12:47 pm

Take a look at CSI (Catalog Search Interface) from the link below.

http://publib.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/DGT2C150/C.0?DT=20060608131553

The REXX code is usually in SYS1.SAMPLIB.

This product can be a real pig to configure to get the exact reports that you need, but for a straight ABC.** it will give you the dataset name and the volume(s) that it resideds on.

Douglas Wilder · Posted: Thu Jul 05, 2007 11:13 pm

Thank You, This works great.