This is hardly definitive, but I've done a quick test, not as quick as it
should be of it, I did a more complete test a few weeks ago, so this is just
for the figures to send.
Browser Win2k/IIS 5.0 WinNT 4.0/IIS 4.0
Netscape 4.7 >15 minutes 6 seconds
IE 5 seconds 5 seconds
The file being viewed uses ssi, there's no noticable slowdown if you're
using a static file. From the results the browser is not the problem since
using IIS 4.0 doesn't display the problem. Someone posted something here
earlier about a client-side fix, I didn't do too much looking into it so I
can't say much about that, but if it does solve the problem it points to a
little of both server/client side. I'm not saying that this is a conspiracy
from MS, but it does appear to be isolated to IIS 5.0 under Win2k,
personally I don't like MS business practices, but Apache still holds the
majority of web servers, and such a practice would only send business their
way.
I would have prefered to not send this to the mailing list, except there
seem to be several people involved in this discussion.
> Oh, I wouldn't start blaming MS for problems in netscape without doing a
tru
> lab test on the issue. I've been programming since the early 1980s, and
I've
> heard some strange conspiracy theories, and the ones about Netscape
running
> on Windows goes into the X-files unless one can truly and independently
test
> the veracity. Just use the best mix of software for your environment, and
> make sure to thoroughly account for the environment in your servlets.
>
> John C.
> -----Original Message-----
> From: Charles Chen [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, October 10, 2000 9:46 AM
> To: [EMAIL PROTECTED]
> Subject: Re: Combination ? [Scanned by Yellow Pages PostMaster]
>
>
> There might be a fastest combination but you will probably never find
it;-)
> In
> my experience, you should never try to find the best combination in
> isolation of
> yuor application requirement in the first place.
>
> IIS + JRun might be a good combination if you are hosting the web from a
NT
> as
> you then get the benefit of using ASP, if the ASP benefit your
application.
> If
> you are hosting your web on a unix platform, IIS would never run!
>
> If you have read the problems earlier on by Matthias Carlsson, Paul
> Copeland,
> and Lietz, Carol about Microsoft, you might want to think about NOT using
> any
> Microsoft software at all;-(
>
>
> Charles
>
>
>
>
>
>
>
> "Christopher K. St. John" <[EMAIL PROTECTED]> on 10/10/2000 01:15:56 PM
>
> To: [EMAIL PROTECTED]
> cc: (bcc: Charles Chen/YellowPages)
> From: "Christopher K. St. John" <[EMAIL PROTECTED]>, 10 October 2000,
> 1:15
> p.m.
>
> Re: Combination ? [Scanned by Yellow Pages PostMaster]
>
>
> Abhay wrote:
> >
> > I have heard that Jrun 3.0 from Allaire with IIS is the best
> combination
> > for servlets to work on. To what extent that's true? If not, then is
there
> > any other combination to suit the need , so as to achieve maximum speed
> and
> > server responce ?
> >
>
> If you're doing a large, expensive site, then there's
> no way to get around doing your own performance testing
> with various combinations of hardware and software. There
> are too many variables to say in advance "this combination
> is fastest", it depends too much on your particular app.
> I understand that this is a very disappointing answer, and
> it is fairly expensive to do that sort of testing, but
> there's really no good way around it.
>
> In any case, chances are that the particular brand of
> servlet engine and web server are going to have much less
> impact than something like the speed of the database, or
> the amount of memory in the server, or how well written
> your servlets are, or which jvm you are using. But maybe
> not...
>
> One thing for sure: the "raw" server speed tests are
> completely useless for predicting real-world performance
> of complex web applications.
>
> Just out of curiosity, was it Alliare or Microsoft that
> told you the JRun/IIS combo was "best"? :-)
>
>
> -cks
>
>
___________________________________________________________________________
> To unsubscribe, send email to [EMAIL PROTECTED] and include in the
body
> of the message "signoff SERVLET-INTEREST".
>
> Archives: http://archives.java.sun.com/archives/servlet-interest.html
> Resources: http://java.sun.com/products/servlet/external-resources.html
> LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
>
>
___________________________________________________________________________
> To unsubscribe, send email to [EMAIL PROTECTED] and include in the
body
> of the message "signoff SERVLET-INTEREST".
>
> Archives: http://archives.java.sun.com/archives/servlet-interest.html
> Resources: http://java.sun.com/products/servlet/external-resources.html
> LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
>
>
___________________________________________________________________________
> To unsubscribe, send email to [EMAIL PROTECTED] and include in the
body
> of the message "signoff SERVLET-INTEREST".
>
> Archives: http://archives.java.sun.com/archives/servlet-interest.html
> Resources: http://java.sun.com/products/servlet/external-resources.html
> LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
>
___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".
Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
