I'm not in the DFSORT team anymore but thought I'd chime in here.  File
size and dynamic allocation is a topic I could  probably spend hours trying
to explain since there are so many nuances.  But to try and answer your
specific questions.

1. When DFSORT can determine the file size, it calculates the required work
space and then spreads that across the number of dynamically allocated work
data sets specified in the DYNALOC installation default or the DYNALLOC run
time override.  Keep in mind, that DFSORT will also determine the amount of
memory object, Hiperspace or Dataspace storage it can use and factor that
into the disk work space calculation.

2. If the programmers specify an estimated file size (FILSZ=Ennn) it will
be ignored if DFSORT is able to calculate the file size on its own.  This
is usually the case for input data sets on disk or managed tape.
But let's say the sort has an OMIT parameter and you know it's going to
exclude most of the records.  You can specify FILSZ=Unnn to tell DFSORT to
use the specified file size even though it's able to calculate it from the
input.

3. Most of the time DFSORT can calculate the file size.  However there are
instances when it cannot.  The most common that come to mind are when the
input is coming from an E15 input exit or unmanaged tape.    In these cases
DFSORT will use what's specified on the FILSZ parameter (whether is Ennn or
Unnn) or you'll get the ICE118I Unknown file size message.



Have a nice day,
Dave Betten
z/OS Performance Specialist - HiPODS
Open Technology and Cloud Performance
IBM Corporation
email:  [email protected]
1-301-240-3809

IBM Mainframe Discussion List <[email protected]> wrote on
09/14/2015 09:59:51 AM:

> From: Walter Marguccio <[email protected]>
> To: [email protected]
> Date: 09/14/2015 10:05 AM
> Subject: DFSORT (old) enhancement
> Sent by: IBM Mainframe Discussion List <[email protected]>
>
> Ciao Massimo,
> I read that excerpt, too, which however refers to unknown file size
> as input.I'd like to know the amount of primary and secondary space
> of dynamicallyallocated SORTWKxx when the input file size is known.
> Which lead me to another query ... 1) I know my programmers specify
> often FILSZ on the PARM2) we have DFSMSrmm3) I'm pretty sure
> BLOCKSET technique is always selected for a SORT operation
> Can I assume from the above that DFSORT is always able to determine
> an accurate file size ? Walter Marguccio
> z/OS Systems Programmer
> BELENUS LOB Informatic GmbH
> Munich - Germany
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to [email protected] with the message: INFO IBM-MAIN
>

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to