IBM Mainframe Forum

we had a request to expand the size of the key field from 15 to 20. Since it is a key, what is the process to expand the field and move data within the segment?

Whilst I'm not an IMS DBA, I believe that the process is about like this:

1. Unload the data base.
2. Modify the DBD appropriately and re-gen it.
3. Modify the unload (I think you'll have to write a program for this, although Mr. Boyo and/or Mr Kolusu may have *Sort tricks for doing it).
4. Delete/define the data bases and secondary indexes (if any).
5. Re-gen the PSBs and ACBs referring to the modified DBD.
6. Cross your fingers and hope that you won't get called out if things go pear-shaped at 0-dark-thirty

Hello and welcome to the forum,

Suggest you ask the DBA if there is already a process in place that you can clone.

The DBA has not done it in a while and was not sure of the process (out original DBA retired). I have saved the segment and the keys to the parents in a dataset so I can increase the key size. What I do with this file is the question? Do I delete hte old and insrt the new or read and replace? I am an IMS programmmer for > 30 years but not DBA

Hello,

I don't "do" ims, but for several other database systems, i:
. unload the original data (both to convert and for a backup)
. reformat the unloaded data to the new format - save this as the initial load (if needed later)
. redefine the database structure wioth the change(s)
. reload the data and ensure all is well

Remember that any "old" backups will not simply reload. If their data is needed, it will need to be reformatted for use (or the old database structure might be available in some test or repair "region").

This is a forum for Beginners and Students. You are welcome here, especially if you provide answers, but you might be more at home with the Sister site, link at the bottom of the page, next to last.