DFSORT Parameter DYNAUTO=IGNWKDD DB2 8.1 in Compatibility Mode

Ale Eba

DFSORT Parameter DYNAUTO=IGNWKDD DB2 8.1 in Compatibility Mode
Hello,
Is anyone using DYNAUTO=YES for DFSORT module ICEAM2? This is recommended by APAR II14047. According to the APAR this parameter will allow dynamic allocation of work datasets even if JCL work datasets are present.

Is there any real benefit of DFSORT using dynamically allocated work datasets compared to JCL defined work datasets? I ran REORG SHRLEVEL NONE for a tablespace with one table and one index. The table has twenty million rows. One run was with SORTDEVT SYSDA SORTNUM 10 and without JCL defined work datasets and the second was with JCL defined work dataset. I didn’t observe any significant difference in CPU time.

Thanks
Ale


---------------------------------
Cheap Talk? Check out Yahoo! Messenger's low PC-to-Phone call rates.

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Mark McCormack

DFSORT Parameter DYNAUTO=IGNWKDD DB2 8.1 in Compatibility Mode
(in response to Ale Eba)
Ale,

If you choose to use explicitly allocated sortworks via dd stmts in jcl,
you must answer two questions. How many sortworks are needed? How big
should they be? If you underallocate them, you may experience B37s and
DB2 utility failure. If you overallocate them, you waste dasd space.
Will using too much space conflict with other concurrently executing
users of that dasd?

If you dynamically allocate the work data sets, you don't need to answer
the second question. DFSORT will determine how big to allocate the work
data sets. I personally find that much more attractive than calculating
what is needed or quessing. I can't remember the last time I used
explicit dd stmts for sortworks. You are not likely to see much
difference in cpu or elapsed time, however, unless you do something that
causes DB2 to do something drastic, like disable index build
parallelism.

My employer is a Syncsort shop, so we never used DFSORT before DB2v8.
We find it useful / necessary with DFSORT to use much larger region
sizes for our DB2 reorg reorg utility steps.

Mark

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm