MVSFORUMS.com

vini · Posted: Tue Feb 24, 2004 2:20 pm Post subject: Discard Header Record

Hi ,

Can we do this with IEBGENER ?
If not , any similar utility which can both Discard the Header (its the first record of a FB file) and Copy remaining detail records ?!! I need Copy because at the same time I have to Club Six 80 byte records into one 480 byte record ..yet discard the first 80 byte record. I know how to do that trick with Blocksize Wink

but hav no clue around discarding of input records. Dont want to change the program.

Thnks.
Vini

P.S : Search on Discard/Remove header did not yield any post ..

Frank Yaeger · Posted: Tue Feb 24, 2004 3:34 pm Post subject:

You can discard the first record and copy the rest of the records with a DFSORT job like this:

vini · Posted: Tue Feb 24, 2004 4:03 pm Post subject:

Frank ,

I got an Error ... checked the description but cannot make out any inconsistency in file dispositions or anything .

Step was as follows :

Frank Yaeger · Posted: Tue Feb 24, 2004 5:01 pm Post subject:

From what you said before, I guess your SORTIN data set has LRECL=80, but you're trying to override it to LRECL=480 to group 6 records. You can't do that in the same step in which you use SKIPREC=1. You need to use SKIPREC=1 first to get rid of the header record and then use LRECL=480 in a second step. Something like this:

vini · Posted: Tue Feb 24, 2004 5:51 pm Post subject:

Frank ,

This time I did not try to do the trick in same Step as Skiprec ...yet I got Strange Errors ...

1) On EXEC IEBGENER in S2.

Informational error : 'Conflicting DCB Parms ..'

This wasnt unexpected but since its Informational I thought the File would have gotten created atleast but I was wrong ..input had a 100 recs and output none Neutral

.

2) On EXEC ICEMAN in S2

CTMC30I CMEM IS ALREADY ACTIVE FOR THIS JOB -

Never seen this error before ..whats it supposed to exactly mean .the description wasnt really helpful Confused

Zero records in output again.

Pls. suggest .. I know this is workable but just not been lucky so far. Smile

Thnks.
Vini

Frank Yaeger · Posted: Tue Feb 24, 2004 6:49 pm Post subject:

My DFSORT job works fine when the number of input records is a multiple of 6 plus 1 (e.g. 61 records). Removing the header record in step 1 results in a multiple of 6 records which fillsl out the LRECL of 480 for step2 evenly. But when the number of records is NOT a multiple of 6 plus 1 (e.g. 100), I get an I/O error for step2. I think to do your LRECL=480 trick successfully, you have to have an even multiple of 6 records.

I don't know what that CTMC30I message means ... I've never seen it.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

vini · Posted: Wed Feb 25, 2004 9:59 am Post subject:

Frank ,

The Actual number of Input records is 8557 inclusive of Header.
100 was at random .. used for example sake. Now 8556 is an even multiple of 6...seems like the problem is elsewhere ...

Here is my version of the JCL.

Frank Yaeger · Posted: Wed Feb 25, 2004 11:29 am Post subject:

I'm confused. You said

vini · Posted: Wed Feb 25, 2004 12:12 pm Post subject:

Frank ,

Sorry for the confusion .. I myself get confused with ICEMAN vs. ICETOOL ..guess I dont understand the difference so messed them up ....to me they probably mean the same thing , guess they arent ..

I was executing ICETOOL ...and after I corrected that to ICEMAN things look better ...sure they do Smile

Theres one thing I am still unable to get to ..its the Clubbing of 6 records of 80 bytes into 1 record of 480 Lrecl...despite doing this in a seperate step and despite using non-temp DS.
Correct me if I am wrong ...to achieve this ..all we need to do is Execute ICEMAN with Copy from a Source DS of 80 Lrecl ,Blksize 480 ..into a 480 Lrecl DS ?!!! and ofcors the Input File should have even multiple of 6 number of records. Both of these things are True in my case ...still ....with ICEMAN I get no error message ..strange ..yet that steps return code is 16.
What I do get in final output are each 80 byte onto a 480 instead of
6 * 80 = 1*480.

Thanks for ur patience.
Vini

Frank Yaeger · Posted: Wed Feb 25, 2004 12:42 pm Post subject:

Vini,

vini · Posted: Wed Feb 25, 2004 2:21 pm Post subject:

Frank ,
Your Awesome !!!
The JCL you provided ..works Perfectly for all my needs ! Very Happy

.

Thanks
Vini

Frank Yaeger · Posted: Wed Feb 25, 2004 3:29 pm Post subject:

Good.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

Frank Yaeger · Posted: Fri Oct 29, 2010 4:03 pm Post subject:

With z/OS DFSORT V1R10 PTF UK90025 or z/OS DFSORT V1R12 PTF UK90026 (Oct,2010), you can now use the new RESIZE operator of DFSORT's ICETOOL to do this more easily and in only one pass like this: