Page 1 of 1

How to compare two ps files having millions of records ?

Posted: Sun Feb 05, 2012 11:08 pm
by nareshv_99
hi,
i have two sequential files having millions of records. how can i compare both the files ?
i tried with x3.13 but it gave space abend. :?

could anyone help me in this ? :? :?

Posted: Mon Feb 06, 2012 10:34 am
by NicC
Use the batch option of 3.13. When the job abends extract the JCL from the spool, amend the space for the listings dataset - or convert it to use a dataset - and resubmit it.

Posted: Mon Feb 06, 2012 10:57 am
by dbzTHEdinosauer
when using the batch option,
i believe you have the option to EDIT the JCL before submitting the job.

you can then copy the jcl to your own 'JCL' pds,
so that you will have a working JOB which you can then edit
and modify for any SUPERC, thus negating the need to go thru the 3.xx panels.

something to keep in mind:
if the record length is larger than 133, you will not have the complete record as output.

so, unless you are using the output (of a million record compare)
as input to something else (and the record length is less than 134)
why don't you use the summary output format (giving you totals)
since i don't know what you are going to do with 100's of thousands,
if not millions of output records.....

Posted: Mon Feb 06, 2012 12:17 pm
by mohan.lal@capgemini.com
:lol:

Posted: Mon Feb 06, 2012 8:16 pm
by nareshv_99
i have got the solution..
i used below ICETOOL toolin

i gave two files on INDD like
//INDD DD DSN=file1,DISP=SHR
// DD DSN=file2,DISP=SHR
//OUTDD DD DSN=file,DIPS=(,CATLG,) ..
..
//TOOLIN DD*
COPY FROM(INDD) to(OUTDD) ON(1,7,P,8,2,BI,10,1,CH) NODUPS
/*

it gave MAX CC=12 as both files are same.. OUTDD dsn is empty..

thnks for you suggestion too.. :)

Posted: Tue Feb 07, 2012 12:07 am
by Frank Yaeger
That syntax is NOT valid for the ICETOOL COPY operator. The cc=12 would be the result of the syntax error. It would have nothing to do with whether the files were the same.

For complete details on correct DFSORT/ICETOOL syntax, see:

http://www.ibm.com/support/docview.wss? ... g3T7000080

Posted: Tue Feb 07, 2012 8:46 am
by nareshv_99
I'm not sure of Syntax.. as i posted it from my home.. i forgot the actual syntax.. i'll check it in office and post it again..

Posted: Wed Feb 08, 2012 12:05 am
by Frank Yaeger
Ok. The syntax for a job to get RC=12 for an empty data set would be quite different (SELECT? COUNT?) so hopefully you got it right.

In the future, please don't post syntax guesses as it just confuses people.

Posted: Wed Feb 08, 2012 10:17 pm
by nareshv_99
nareshv_99 wrote: //TOOLIN DD*
COPY FROM(INDD) to(OUTDD) ON(1,7,P,8,2,BI,10,1,CH) NODUPS
/*

it gave MAX CC=12 as both files are same.. OUTDD dsn is empty..

thnks for you suggestion too.. :)
it was not COPY.. it is SELECT..
//TOOLIN DD*
SELECT FROM(INDD) to(OUTDD) ON(1,7,P,8,2,BI,10,1,CH) NODUPS
/*
Frank, sorry for posting wrong syntax..i can't post anything from my office thats why i posted what i remembered..

Posted: Wed Feb 08, 2012 11:10 pm
by Frank Yaeger
That syntax is still invalid - it should be:

SELECT FROM(INDD) TO(OUTDD) ON(1,7,PD) ON(8,2,BI) ON(10,1,CH) NODUPS
Frank, sorry for posting wrong syntax..i can't post anything from my office thats why i posted what i remembered.
All of the DFSORT books are available at:

http://www.ibm.com/support/docview.wss? ... g3T7000080

So you could look up the correct syntax even from home.