IRX0662E error whiel reading flat filein REXX

mosinjamadar · New User Joined: 26 Sep 2007 Posts: 42 Location: pune

Hi all

I am getting IRX0662E error when reading a flat file which contains 10,000,000 records. Below is peace of code which i am using to read the file

"EXECIO * DISKR INDD (STEM NEWVAR."

error is
NO OF RECORDS
IRX0662E EXECIO error. Unable to obtain storage.

please help me out with any other way I can over come this

thanks

enrico-sorichetti · Posted: Mon Apr 06, 2009 6:00 pm

already debated to death,
the EXCIO as coded will read into storage all the dataset records
for large number of records that' s the least You can expect

process them in chunks , search the forum for samples

another suggestion ...
review Your process and evaluate alternatives
for large datasets processing REXX is not the best tool around

expat · Posted: Mon Apr 06, 2009 6:05 pm

dick scherrer · Posted: Mon Apr 06, 2009 8:48 pm

Hello,

bala kalluri · New User Joined: 08 Jan 2009 Posts: 21 Location: INDIA

Even I'm facing the same problem, rexx isn't able to handle huge amounts of data.

Could you tell me, what alternate way have you chosen.

dick scherrer · Posted: Wed Apr 15, 2009 9:21 pm

Hello,

Depending on what you want to do, you could use real code or possibly the sort product that is in use on your system.

bala kalluri · New User Joined: 08 Jan 2009 Posts: 21 Location: INDIA

HI All,

I have huge number of data in a flat file.
My requirement is to find a particular string in the file.
IF i implement the logic of reading record one by one and searching for a particulart string, then its throwing "storage exhaust" problem.

If not rexx, is there any program in cobol which can help me out to find the record number which contains the searched string details.

Thanks,
Bala

superk · Posted: Wed Nov 25, 2009 5:10 pm

dbzTHEdinosauer · Posted: Wed Nov 25, 2009 5:34 pm

or the 3 dot option searchfor?

expat · Posted: Wed Nov 25, 2009 5:37 pm

Do you really mean reading "one by one" or do you mean trying to load a stem with the datset.

This has been discussed numerous times, and you can process the file into a stem in manageable segements, or you could (and most probably should) use your sort product to do this.

dick scherrer · Posted: Wed Nov 25, 2009 9:35 pm

Probably, i oversimplfy, but ten million records should not be processed in rexx. . .

d

bala kalluri · New User Joined: 08 Jan 2009 Posts: 21 Location: INDIA

I want read record line by line and to capture the one which satisfies my search criteria. I have huge number of records in range of 10-20 lakhs.
Reading records in stem variable will not help, it will throw storage exhausted error.

I cant use sort utility here, all records contains different data.

expat · Posted: Mon Nov 30, 2009 7:49 pm

Please elaborate why you can not use sort to select only those records which contain the required string. Using the SS capabilities of DFSORT should not cause you any problems.

RECFM & LRECL of the input file, and possible positions of the required string.

dbzTHEdinosauer · Posted: Mon Nov 30, 2009 7:54 pm