Job Abends with a file status code of 24 when writing VSAM

madhav_in · New User Joined: 21 Nov 2006 Posts: 12 Location: chennai

Hi,
I have been getting this abend for quite sometime in the batch processing. The job uses various VSAM files in the I/O mode and sometimes the job abends when trying to write into a particular file with a return code of 24. I checked the forum about this error and found that increasing the secondary extent could do the job. But what baffles me is that , Since the files are I/O files , I need to restore them from backups before restarting the job. Now when I restore these I/O files ( without making anyother changes ) and restart, the job runs fine without any problem. Is there a logical reason to this type of an abend or is it just some environment issue ?

dbzTHEdinosauer · Posted: Tue Jul 29, 2008 1:14 pm

When you restore them, haven't you effectively REORG'd them?

expat · Posted: Tue Jul 29, 2008 1:29 pm

madhav_in · New User Joined: 21 Nov 2006 Posts: 12 Location: chennai

Thanks a lot for your reply

, I dont understand what this REORG means ? How does restoring the file cause this REORG, I am just going to copy the backup file to the VSAM file, this causes REORG ? Is there a way to tackle this issue from my side ?

expat · Posted: Tue Jul 29, 2008 6:48 pm

A good intro to VSAM

Have a read, it is well worth the effort.

But ............. a restore from backup is a little like an initial load of the VSAM cluster. Every record is in its correct place with no wasted internal space. Once a file is written to and has CI/CA splits then some space may be 'Free' but is unusable.

madhav_in · New User Joined: 21 Nov 2006 Posts: 12 Location: chennai

Hi Expat I will surely go through the IBM redbooks link that you had mentioned to understand why this happens. So as u said when i copy from a backup, it is like an initial load and that any wasted space will be utilized. Then will having an IDCAMS step copying the backup file to the original VSAM file before the program runs, do the trick of solving the problem ?

expat · Posted: Wed Jul 30, 2008 11:32 am

It will reorg the cluster, but as nI said about 8 posts ago, have a word with the people who maintain the cluster and get the space allocations changed to allow for contingency.