Date: Fri, 26 May 2006 14:22:49 +0000
Reply-To: toby dunn <tobydunn@HOTMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: toby dunn <tobydunn@HOTMAIL.COM>
Subject: Re: Comparison of Proc Syncsort and SAS 9.1.3 Sort on IBM Z/os
systems
In-Reply-To: <200605261414.k4QAkSpH032625@mailgw.cc.uga.edu>
Content-Type: text/plain; format=flowed
Paul .
Well I for one miss your adroitness. Other than the ussual fare of
questions spanning the ussualy gamet all you have missed as of late is Ian
kicking my tailend around the block over some macro quoting philospophical
issues.
Toby Dunn
From: "Paul M. Dorfman" <sashole@BELLSOUTH.NET>
Reply-To: "Paul M. Dorfman" <sashole@BELLSOUTH.NET>
To: SAS-L@LISTSERV.UGA.EDU
Subject: Re: Comparison of Proc Syncsort and SAS 9.1.3 Sort on IBM Z/os
systems
Date: Fri, 26 May 2006 10:14:33 -0400
David,
I am not sure this Paul is still the prime authority on this one...
methinks Paul Choate has been doing lots of research in this direction
lately. I will be able to do an independent real-world comparison in a
short order, though, being on a short gig at a yet another z/OS
installation. One of the first thing I have checked into was the
availability of the core SyncSort and PROC SYNCSORT in particular. The
most recent core is present, however PROC SYNCSORT is outdated and has to
be updated and installed to get it up and running with SAS 9.1.3. This
should be done in a couple of weeks, I am being told; then we will see.
I concur with your doubts about native SAS sort matching PROC SYNCSORT
performance under z/OS ever, but this does not exhaust the advantage. It
should be born in mind that PROC SYNCSORT has the [undocumented] SUM
statement behind its belt, which has so far outperformed an equivalent
NWAY MEANS on summation at the ratio of mere 10:1. And I believe that Paul
Choate has now discovered that the [also undocumented] MIN and MAX
statements are now active in the same sense. I have never seen those work,
so I am looking forward to test them in a near future when the sortfare
has been updated and while I have an access to the Real Computer, and will
report the results.
To all my friends at sas-l I would like to say that I miss more vigorous
interaction with the list on my part greatly, yet my other duties have
currently kept me from even reading the list in full, let alone exhibiting
but a tiny fraction of my former perceived adroitness. However, I hope to
return, at least partially, to the normal modus operandi, once the
activities keeping me from it have sufficiently subsided.
Kind regards
------------
Paul Dorfman
Jax, FL
------------
On Thu, 25 May 2006 22:27:07 -0700, David L Cassell
<davidlcassell@MSN.COM> wrote:
>SRhoades@US.IMSHEALTH.COM wrote:
>>Has anyone run tests of efficiencies between Proc Syncsort and SAS Proc =
>>Sort on the IBM Z/os system and version 9.1.3 of SAS that they would =
>>share? We are being told that the speeds are now equivalent.
>>=20
>>Thanks;
>>=20
>>Steve Rhoades=20
>
>I believe that Paul Dorfman has written on this subject in SAS-L enough
>times that he ought to be consulted on this one. (Sorry, but at this
point
>I'm entertaining myself seeing a midget in a bellhop outfit yelling "CALL
>FOR PAUL DORRRRRRRFMAN!!" :-)
>
>But I would be pretty darn impressed if SAS was now able to match
>the speed of SyncSort on z/OS. I think SAS is as fast as SyncSort on
>unix, but not on z/OS. Just get both, and then let SAS use SyncSort
>as its SORTPGM= option.
>
>z/OS is Syncsort's home field, and they've watered down the basepaths
>to keep other teams from going as fast. :-) :-)
>
>David
>--
>David L. Cassell
>mathematical statistician
>Design Pathways
>3115 NW Norwood Pl.
>Corvallis OR 97330
>
>_________________________________________________________________
>Is your PC infected? Get a FREE online computer virus scan from McAfeeŽ
>Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963
| 
