MVSFORUMS.com

NASCAR9

I have searched so don't bury if my searches didn't turn up what I need. Embarassed

I need a way to grab all files with the same high level qualifier and concatinate them into 1 dataset. The number of files is unknown. The record length is fixed all records same size. My choices for tools are IBM utility or Synsort.

I would apperciate any help.
_________________
Thanks,
NASCAR9

CICS Guy · Intermediate Joined: 30 Apr 2007 Posts: 292 Topics: 3

Up to the concatination limit, just concatinate them....

NASCAR9 · Posted: Thu Sep 27, 2007 2:01 pm Post subject:

CICS Guy, thanks for your reply! I like your reply it sounds easy, but I'm unclear how or what concatination limit your taking about. Embarassed

Do you have an example?
_________________
Thanks,
NASCAR9

CICS Guy · Intermediate Joined: 30 Apr 2007 Posts: 292 Topics: 3

I'm sure there is a (JCL?) manual that specifies the max number of datasets that can be concatenated....

semigeezer · Posted: Thu Sep 27, 2007 6:21 pm Post subject:

Hopefully whatever person asked you to do this is not really concerned with how it gets done because neither of those solutions will work for a large number of data sets.

Since there are limits on concatenation and on the # of steps in a job, I'd suggest that you write a small program to get the data set names, allocate the target with disp=mod and copy each data set's contents to the output data set. This should probably be about 20 lines in Rexx, including error checking. Find the data set names with LMDLIST, do the allocation with LMINIT (and LMFREE) and do the copying with LMCOPY. Examples of all of these abound on the web. One thing to note is you may need to do an LMINIT/LMFREE on the target data set for each copy rather than doing it just once (due to an ISPF limitation).

vkphani · Intermediate Joined: 05 Sep 2003 Posts: 483 Topics: 48

NASCAR9,

Get all the file names by goving your high level qualifier and then concatenate them in the job.

videlord · Beginner Joined: 09 Dec 2004 Posts: 147 Topics: 19

1. ISPF 3.4, list the data set whtat you want
SAVE command, save to a data set A
(or you can use LISTDS)
2. EDIT data set, to be your JCL DD card
(DFSORT or IEB* utility can be used)

It can be done in batch mode too, or you can code a simple REXX routine

hchinnam · Beginner Joined: 04 Feb 2005 Posts: 31 Topics: 10

There is a utility which does something similar to what you are asking.

We were using it for taking the BKUP. It takes the wild card as input and copies all the datasets into a single output.

(i.e you can copy HLQ.* to some dataset name)

BUT

It will be a solution if your requirement is only to take BKUP.
Reason being it stores the data in it's own format. And you need to
execute the same program again to retrieve any of the data.

If you are interested in using this, the name goes some thing like ADDR**
(some address utility).

vkphani · Intermediate Joined: 05 Sep 2003 Posts: 483 Topics: 48

It is ADRDSSU. You can check the below link for example.
ADRDSSU Documentation

NASCAR9 · Posted: Fri Sep 28, 2007 10:10 am Post subject:

Thank you all for your responses! Very Happy

I will investigate all options.
_________________
Thanks,
NASCAR9

NASCAR9 · Posted: Fri Sep 28, 2007 11:48 am Post subject:

I would like to add some clarification to my post. I'm not doing a backup.
The files I'm processing are for an outside vendor who will FTP files with a HLQ of PRXSOL.*
This is the only part of the file name I can require.
We created a REXX program. It seems to work ok, just maybe a little more tweaking.
I will post it when were done.
_________________
Thanks,
NASCAR9

NASCAR9 · Posted: Tue Nov 13, 2007 1:33 pm Post subject:

An update to this post. The requirments have changed and now we will be Pulling the files, not having them pushed to us. The lan support group will be responsible for this activity. They will create a single file that will be FTP'ed to the mainframe and processed.
Thanks Again! Very Happy

_________________
Thanks,
NASCAR9