Have you used FILEMON to see what the addition I/O is?
----- Original Message -----
From: "Albert Berry" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, June 26, 2001 11:50 AM
Subject: Re: RE: More SET MULTI vs speed Tests
> This test has been run by us on a number of servers, on site and at client
sites with similar results. RBTI has been able to replicate the problem on
their servers. The only issue is more than one user on a W2K server.
>
> [EMAIL PROTECTED] wrote:
> >
> > What kind of server did you test on?
> >
> > Ben Petersen
> >
> >
> > On 26 Jun 2001, at 9:52, Doug Daily wrote:
> >
> > >
> > > We just ran a test on a server and across lines we have
> > > zero control of, and are very unhappy with.
> > >
> > > With Multi ON, staticdb OFF Multi ON, Staticdb ON
> > >
> > > one user 15 sec 15 sec
> > >
> > > two users 199 sec ! 189 sec
> > >
> > >
> > > We're going to test on another server where we have a bit
> > > more control of the environment.
> > >
> > > But unfortunately we're probably stuck with the first
> > > situation, which means we're doomed. No fix.
> > >
> > > Douglas Daily
> > > Payroll Manager
> > > Facilities Services, box 352215
> > > 206-221-4357
> > >
> > >
> > > On Tue, 26 Jun 2001, Albert Berry wrote:
> > >
> > > > We tested with Staticdb on and off - it makes no difference. The
only factor that changes the t
> > imes dramatically like this is whether or not a second user is connected
to the db. Multi on/off, s
> > taticdb on/off with and without fastlocks - only significant difference
is caused by connecting a s
> > econd user - who need do absolutely nothing, just connect, to kill the
performance. W2000 servers
> > only.
> > > >
> > > > There was a note on the list yesterday about server settings that
our network team will have a
> > look at.
> > > >
> > > >
> > >
> > >
> >
> >
> >
> __________________________________________________________________
> Get your own FREE, personal Netscape Webmail account today at
http://webmail.netscape.com/
>
