|Date: ||Mon, 5 May 1997 15:52:32 -0400|
|Reply-To: ||Laurie Fleming <lfleming@ACTRIX.GEN.NZ>|
|Sender: ||"SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>|
|From: ||Laurie Fleming <lfleming@ACTRIX.GEN.NZ>|
|Organization: ||Actrix Networks Limited|
|Subject: ||Re: 600 invocations of sas 6.12?? -reply|
> Original question:
> > I just recently heard a rumor that sas 6.12 for windows produces
> > problems
> > for one's computer after the 600th invocation. Is this urban legend?
> > Uninformed rumor? Real problem?
> > Reply.
> > The Alert Notes that come with the 6.12 installation software state:
> > "Repeated invocations of Release 6.12 of the SAS System under
> > Microsoft Windows 95 will result in a memory leak that will cause the
> > system to malfunction. It has been determined that about 600 invocations
> > of the SAS System in interactive mode and about 2000 invocations in
> > non-interactive mode will reproduce this problem."
> > I take this to mean one SAS session invoking another - I hope it is not
> > each time one starts SAS.
> > Carman Pagano
> > Quaker Chemical Corporation
> > Conshohocken, PA 19428 USA
> I would strongly guess that this refers to separate invocations in the
> same windows session. In my experience, restarting windows has always
> recovered memory lost by leaky applications.
> Jack Ostroff email@example.com
> Pfizer Central Research, Groton, CT
Don Stanley and I have confirmation that the more data steps and macros called
session (especially lots of %do loops, which I specialise in (^-8 ), make each
invocation slower and slower - in the space of an afternoon, a run time over
100,000 observations slipped from 9.1 seconds to around 24 seconds. Getting out
of SAS (this is on a fast Sparc box) and back in again returned it to the
It's a bloody nuisance.
Laurie Fleming | On a clear disk, | (+64 4) 479-1589
4 Kenya St | you can seek forever. | (+64 21) 688-140
Ngaio | | firstname.lastname@example.org
Wellington 6004 | | email@example.com
New Zealand | |