You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by "Dietz, Phil E." <PE...@West.com> on 1999/05/19 19:06:44 UTC

apache benchmarks in general

All of the Mindcraft articles, in my mind, make Apache look at fault -- when
in reality it's the Linux scheduler.  

Another impression appears that no one knows how to configure Apache for
performance nor know which OS Apache would run on best.  People guess Linux
since it's the fad OS nowadays.
That indicates people want docs telling them what to use.

I propose that the Apache group benchmark the leading OS's and provide the
results in a document for the Apache distribution.

Then Apache can not be inadvertently-defamed by tests like Mindcraft.

If people see that the Gameboy version of Apache is slow then there's no
reason to "commision Mindcraft to benchmark" anything or see articles flying
across the Internet preaching Apache stinks.


Re: apache benchmarks in general

Posted by Dean Gaudet <dg...@arctic.org>.

On Wed, 19 May 1999, Dietz, Phil E. wrote:

> All of the Mindcraft articles, in my mind, make Apache look at fault -- when
> in reality it's the Linux scheduler.  

Yes and no. 

Apache is not the world's fastest webserver. 

> Another impression appears that no one knows how to configure Apache for
> performance nor know which OS Apache would run on best.  People guess Linux
> since it's the fad OS nowadays.
> That indicates people want docs telling them what to use.

No, they didn't guess linux -- that was deliberate.  The linux/apache
combo is being targetted for *marketing* reasons.  Documentation is
pointless -- there's already docs telling people how to configure apache,
and there's a highperformance.conf file ... but few of these "reviews"
actually use that available information.

If they had wanted to just contrast linux scalability with NT scalability
they would have used zeus or thttpd or something else.  No, they wanted to
show a popular combination against its popular competitor.

Yeah it hurts, but apache just isn't the fastest at the benchmarks people
are using.  It still seems to do well in the real world.  But whatever,
the real world has nothing to do with benchmarks. 

Dean