httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Timothy Stone <>
Subject Re: [users@httpd] Apache 2.0.43 rather slow in Windows
Date Wed, 30 Jul 2003 18:15:11 GMT
Joshua touched on several points here that deserve expansion upon:

> It depends entirely on the type of dynamic file.  In general, IIS is going
> to be faster than apache since it is essentially a kernel-embedded HTTP
> server.  But for many (if not most) types of content and networks, the
> difference will be irrelevant since the limiting factors will be network
> and dynamic content generation speed.

Most dynamic developers, esp. JSP/Servlet, will tell you that Apache is 
'da bomb' for static delivery. Out of the box, Apache *will not* serve 
dynamic content, any dynamic content, relying on modules to serve PERL, 
PHP and the like. Apache must be connected to a Servlet/JSP container, 
i.e. Tomcat, for Java delivery, and the variations go on.

Additionally, as Joshua points out, your benchmarking is flawed and 
unfair. Because IIS is essentially a kernel embedded HTTP server, the 
performance marks are going to skew heavily in IIS's favor.

Steven J. Vaughan-Nichols points out in his /Shutdown/ piece for this 
month's /Linux Magazine/, "Server 2003 does, however, have two 
advantages over enterprise-quality Linux distros like RHES and 
UnitedLinux. It's much faster at both file and web page delivery. No, 
that's not FUD, those are the facts."

SJVN cites the many reasons why, but the gist is that IIS (in W2K and 
W2k3 is assisted by, or part of, the kernel. It's just going to be 
ass-faster than Apache because Apache isn't assisted by tricks in the 
kernel. SJNV points out TUX, a kernel-assisted HTTP server runs 
neck-and-neck with IIS. SJVN's article isn't going to be public for 3 
months :( /L-M/ policy to hook circulation and subscribers. But it's 
commentary not HOW-TO anyway.


The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:> for more info.
To unsubscribe, e-mail:
   "   from the digest:
For additional commands, e-mail:

View raw message