You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by "Jason A. Dour" <ja...@dour.org> on 1998/06/15 19:26:43 UTC
Apache In The News: Latest PC Week Article
-----BEGIN PGP SIGNED MESSAGE-----
Hello all,
Mr. Rapoza originally called me before emailing the group about
performance tuning for hte latest PC Week article on Apache. I just spoke
with him, and he gave me the address of the article, for those who wish to
read the latest piece of PC-Weekisms...
http://www.zdnet.com/pcweek/reviews/0615/15apache.html
Enjoy...or try to...
Jason
# "Jason A. Dour" <ja...@dour.org> (http://dour.org/jason/)
# Finger for URLs, PGP Key, Geek Code, PJ Harvey info, et cetera.
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBNYVZWJo1JaC71RLxAQHWVgP/Y7zut1fo1/4vTRLgRvL7SoZ0kRwWo5dW
johVsjZk1UO3k2zdByrnzPk3yKYnHHXdfUeKC4Y62NVCgyYhEnK04sUghsXN5LU9
NlMv2GzYws1DAQ0rchXwRSzF/g71ZUicfC7UDCJ200QLrxzgINvjeJ/loYgwOajC
yInnBwtgmj8=
=/If7
-----END PGP SIGNATURE-----
Re: Apache In The News: Latest PC Week Article
Posted by Dean Gaudet <dg...@arctic.org>.
The data set was only 60Mb when they had me in to tune it for a previous
article... should have been the same dataset you saw too... webbench 2.0
or something like that?
Dean
On Mon, 15 Jun 1998, Marc Slemko wrote:
> Your 500 requets/sec qare a _lot_ easier than theirs probably.
>
> This is over a large dataset (at least a few hundred megs) of varying
> sizes.
>
> On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
>
> > > http://www.zdnet.com/pcweek/reviews/0615/15apache.html
> > >
> > > Enjoy...or try to...
> >
> > Ouch! I think what we really need is to put together a reference
> > configuration on a couple of different operating systems to give benchmark
> > people a place to start. They have Apache topping out at 500 requests per
> > second and IIS levelling off at 1000 requests per second. In testing
> > mod_php I have easily gone above 500 requests per second, and that is with
> > the PHP module. On Solaris though, and not Linux.
> >
> > -Rasmus
> >
>
>
Re: Apache In The News: Latest PC Week Article
Posted by Marc Slemko <ma...@worldgate.com>.
IIS will generally beat Apache by a fair bit on benchmarks consisting of
static content due to the extra caching done just for that.
CGIs, dynamic content, long term stability, etc. is a whole different
issue.
On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
> > Your 500 requets/sec qare a _lot_ easier than theirs probably.
> >
> > This is over a large dataset (at least a few hundred megs) of varying
> > sizes.
>
> Could be. They don't expand on that aspect of it. If IIS was twice as
> fast as Apache on similar hardware, then I don't think I would be getting
> all these NT/IIS/ASP to Unix/Apache/PHP converts showing up on the PHP3
> mailing list. Most of them cite performance and stability as their reason
> for switching.
>
> -Rasmus
>
Re: Apache In The News: Latest PC Week Article
Posted by Chris Tacy <ch...@enginered.com>.
also de-cloaking...
i don't know about "benchmark optimization" but IIS is *seriously*
optimized for serving static files. our benchmarks (Compaq 4way) showed
pretty outrageous performance for this sort of usage. of course... throw
in authentication or error handling or dynamic content (of any sort) and
all bets are off.
-c
Ian Kallen wrote:
>
> Temporarily delurking here...
> I did some benchmarking recently and was surprised at how high IIS and
> Enterprise Server came out. Just like MS and Sun torque their JVM's to do
> well on the CaffeineMark benchmarks, I suspect they similarly optimize to
> handle the extremely artificial environment of benchmarks. For the
> article I wrote up, I emphasized specifically that benchmarks measure one
> thing and it's one thing that is not very important to me: how well the
> server handles benchmark loads. If folks here are really worried about
> this then developing algorithms that optimize handling large numbers of
> GETs from a proportionately small number of clients on low latency
> networks should be a priority. Otherwise, take it for what it is:
> measurements that have little bearing on real world performance. It's not
> that I don't think the benchmark authors are sincere, I just don't think
> they really understand the number of crazy factors that effect real world
> performance: network latency, buggy browsers and goofy tcp/ip
> implementations are not simulated.
> -Ian
>
> On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
> :> Your 500 requets/sec qare a _lot_ easier than theirs probably.
> :>
> :> This is over a large dataset (at least a few hundred megs) of varying
> :> sizes.
> :
> :Could be. They don't expand on that aspect of it. If IIS was twice as
> :fast as Apache on similar hardware, then I don't think I would be getting
> :all these NT/IIS/ASP to Unix/Apache/PHP converts showing up on the PHP3
> :mailing list. Most of them cite performance and stability as their reason
> :for switching.
>
> --
> Where there's no emotion, there's no motive for violence.
> -- Spock, "Dagger of the Mind", stardate 2715.1
--
#################################################
chris tacy president and co-founder
fire engine red http://www.enginered.com/
Re: Apache In The News: Latest PC Week Article
Posted by Dean Gaudet <dg...@arctic.org>.
Yup that's one of my motivations for the flow stuff... (but I want to do
something that helps real world sites too).
Dean
On Mon, 15 Jun 1998, Ian Kallen wrote:
>
> Temporarily delurking here...
> I did some benchmarking recently and was surprised at how high IIS and
> Enterprise Server came out. Just like MS and Sun torque their JVM's to do
> well on the CaffeineMark benchmarks, I suspect they similarly optimize to
> handle the extremely artificial environment of benchmarks. For the
> article I wrote up, I emphasized specifically that benchmarks measure one
> thing and it's one thing that is not very important to me: how well the
> server handles benchmark loads. If folks here are really worried about
> this then developing algorithms that optimize handling large numbers of
> GETs from a proportionately small number of clients on low latency
> networks should be a priority. Otherwise, take it for what it is:
> measurements that have little bearing on real world performance. It's not
> that I don't think the benchmark authors are sincere, I just don't think
> they really understand the number of crazy factors that effect real world
> performance: network latency, buggy browsers and goofy tcp/ip
> implementations are not simulated.
> -Ian
>
> On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
> :> Your 500 requets/sec qare a _lot_ easier than theirs probably.
> :>
> :> This is over a large dataset (at least a few hundred megs) of varying
> :> sizes.
> :
> :Could be. They don't expand on that aspect of it. If IIS was twice as
> :fast as Apache on similar hardware, then I don't think I would be getting
> :all these NT/IIS/ASP to Unix/Apache/PHP converts showing up on the PHP3
> :mailing list. Most of them cite performance and stability as their reason
> :for switching.
>
> --
> Where there's no emotion, there's no motive for violence.
> -- Spock, "Dagger of the Mind", stardate 2715.1
>
>
Re: Apache In The News: Latest PC Week Article
Posted by Ian Kallen <ia...@gamespot.com>.
Temporarily delurking here...
I did some benchmarking recently and was surprised at how high IIS and
Enterprise Server came out. Just like MS and Sun torque their JVM's to do
well on the CaffeineMark benchmarks, I suspect they similarly optimize to
handle the extremely artificial environment of benchmarks. For the
article I wrote up, I emphasized specifically that benchmarks measure one
thing and it's one thing that is not very important to me: how well the
server handles benchmark loads. If folks here are really worried about
this then developing algorithms that optimize handling large numbers of
GETs from a proportionately small number of clients on low latency
networks should be a priority. Otherwise, take it for what it is:
measurements that have little bearing on real world performance. It's not
that I don't think the benchmark authors are sincere, I just don't think
they really understand the number of crazy factors that effect real world
performance: network latency, buggy browsers and goofy tcp/ip
implementations are not simulated.
-Ian
On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
:> Your 500 requets/sec qare a _lot_ easier than theirs probably.
:>
:> This is over a large dataset (at least a few hundred megs) of varying
:> sizes.
:
:Could be. They don't expand on that aspect of it. If IIS was twice as
:fast as Apache on similar hardware, then I don't think I would be getting
:all these NT/IIS/ASP to Unix/Apache/PHP converts showing up on the PHP3
:mailing list. Most of them cite performance and stability as their reason
:for switching.
--
Where there's no emotion, there's no motive for violence.
-- Spock, "Dagger of the Mind", stardate 2715.1
Re: Apache In The News: Latest PC Week Article
Posted by Rasmus Lerdorf <ra...@lerdorf.on.ca>.
> Your 500 requets/sec qare a _lot_ easier than theirs probably.
>
> This is over a large dataset (at least a few hundred megs) of varying
> sizes.
Could be. They don't expand on that aspect of it. If IIS was twice as
fast as Apache on similar hardware, then I don't think I would be getting
all these NT/IIS/ASP to Unix/Apache/PHP converts showing up on the PHP3
mailing list. Most of them cite performance and stability as their reason
for switching.
-Rasmus
Re: Apache In The News: Latest PC Week Article
Posted by Marc Slemko <ma...@worldgate.com>.
Your 500 requets/sec qare a _lot_ easier than theirs probably.
This is over a large dataset (at least a few hundred megs) of varying
sizes.
On Mon, 15 Jun 1998, Rasmus Lerdorf wrote:
> > http://www.zdnet.com/pcweek/reviews/0615/15apache.html
> >
> > Enjoy...or try to...
>
> Ouch! I think what we really need is to put together a reference
> configuration on a couple of different operating systems to give benchmark
> people a place to start. They have Apache topping out at 500 requests per
> second and IIS levelling off at 1000 requests per second. In testing
> mod_php I have easily gone above 500 requests per second, and that is with
> the PHP module. On Solaris though, and not Linux.
>
> -Rasmus
>
Re: Apache In The News: Latest PC Week Article
Posted by Rasmus Lerdorf <ra...@lerdorf.on.ca>.
> http://www.zdnet.com/pcweek/reviews/0615/15apache.html
>
> Enjoy...or try to...
Ouch! I think what we really need is to put together a reference
configuration on a couple of different operating systems to give benchmark
people a place to start. They have Apache topping out at 500 requests per
second and IIS levelling off at 1000 requests per second. In testing
mod_php I have easily gone above 500 requests per second, and that is with
the PHP module. On Solaris though, and not Linux.
-Rasmus
Re: Apache In The News: Latest PC Week Article
Posted by "Jason A. Dour" <ja...@dour.org>.
-----BEGIN PGP SIGNED MESSAGE-----
On Mon, 15 Jun 1998, Marc Slemko wrote:
> Sigh. He had to have it done like the day or perhaps the day after he
> mailed the list, so when I responded he said "oops, too bad."
Yeah, well, that was *sorta* my fault. He called me MANY days in advance
and asked me about it. I had to archive the voice mail and get to it
later because I've been so swamped lately.
I wasn't able to refer him to emailing the group until *today*. Don't
EVEN ask me how he decided I was the performance tuning specialist of
Apache...maybe it was that letter I sent in on behalf on the AG that got
me listed as "Apache Contact" or something weird. *sigh* Sorry guys...I
sorta dropped the ball on this one.
Jason
# "Jason A. Dour" <ja...@dour.org> (http://dour.org/jason/)
# Finger for URLs, PGP Key, Geek Code, PJ Harvey info, et cetera.
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBNYV/Opo1JaC71RLxAQGCrwP/bwCgMnCghJcYZ+tId3q9nxvHXoXFxmyr
OM4e5mNncpN98sh/tLxJJEy0Zud57f6AAEvvd0udKXzzareQ0LJ8m3yUvQUTpco+
FR5+KYscK3aJOw5jMFo8L+j7UgDm5lY0mPz4jqq9hXpKpEDaKY4iNFEExxi6cLSk
C10R++mYjcE=
=rFGv
-----END PGP SIGNATURE-----
Re: Apache In The News: Latest PC Week Article
Posted by Marc Slemko <ma...@worldgate.com>.
Sigh. He had to have it done like the day or perhaps the day after he
mailed the list, so when I responded he said "oops, too bad."
Doesn't test well my ass.
They are using a different OS but can magically attribute the difference
in performance to Apache.
To be fair, he did say he would have preferred not to benchmark at all
this time but he didn't have a choice.
On Mon, 15 Jun 1998, Jason A. Dour wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
>
> Hello all,
>
> Mr. Rapoza originally called me before emailing the group about
> performance tuning for hte latest PC Week article on Apache. I just spoke
> with him, and he gave me the address of the article, for those who wish to
> read the latest piece of PC-Weekisms...
>
> http://www.zdnet.com/pcweek/reviews/0615/15apache.html
>
> Enjoy...or try to...
>
>
>
> Jason
> # "Jason A. Dour" <ja...@dour.org> (http://dour.org/jason/)
> # Finger for URLs, PGP Key, Geek Code, PJ Harvey info, et cetera.
>
> -----BEGIN PGP SIGNATURE-----
> Version: 2.6.2
>
> iQCVAwUBNYVZWJo1JaC71RLxAQHWVgP/Y7zut1fo1/4vTRLgRvL7SoZ0kRwWo5dW
> johVsjZk1UO3k2zdByrnzPk3yKYnHHXdfUeKC4Y62NVCgyYhEnK04sUghsXN5LU9
> NlMv2GzYws1DAQ0rchXwRSzF/g71ZUicfC7UDCJ200QLrxzgINvjeJ/loYgwOajC
> yInnBwtgmj8=
> =/If7
> -----END PGP SIGNATURE-----
>