Re: XML Configuration [Was: Re: Caucho faster than mod_perl?]

[Perrin Harkins]

Matt Sergeant wrote:
> Also, what's different between Resin's smart caching and mod_perl's? Is it
> just like StatINC?

It can cache output in a way similar to Mason.  They call this "Smart
Caching".

It reloads servlet classes pretty cleanly when code is updated, but
doesn't track dependencies, so it may be not as good as StatINC there.

It also does a very nice job with JSP pages, turning them into
first-class servlets.  This produces less overhead than the current crop
of perl-in-HTML tools available for mod_perl.  We tested the same "Hello
World" app as a servlet and as a JSP page and found no difference in
performance.

- Perrin

Re: XML Configuration [Was: Re: Caucho faster than mod_perl?]

[Sean Chittenden]

Personally, I don't care.  A conf file is a conf file.  From the
advocacy standpoint, I think it's probably a smart move on the part of
mod_perl/httpd developers.  There is a HUGE value in hyped up techologies with
upper management.  I wish I were kidding when I say that in the past I have
seen several managers choose utilities because they have XML configurations (or
some swift GUI, etc.).  The product, in four of the six instances, turned out
to be a nightmarish hell that entailed many sleepless nights on the phone with
the Exodus NOCies talking them through how to coach servers back to life
(mod_perl to the rescue!!!).
Anyway, I think this would be pretty groovin' and
would lend itself to a "public friendly" system that would sell better, which,
seems to be mod_perls big problem (IMHO).  Programmers are doing the
marketing/advertising, etc...  managers, who make the decissions listen to
non-techies (typically) and are going w/ Java. 
The question remains then, what
part of the config outside of the httpd.conf could be put into an XML config?

--SC

On Wed, 02 Feb 2000, Matt Sergeant wrote:
> On Wed, 02 Feb 2000, Perrin Harkins wrote:
> > > I'm not trying to belittle Resin in any way -- in fact I'm impressed by
> > > both its design and performance.  In learning about Resin during these
> > > tests, I found that JSP is in many ways easier to use than mod_perl. The
> > > smart caching that Resin does (with compiling .java -> .class, etc.) is
> > > effective, and the XML configuration is a joy to deal with compared to
> > > Apache's httpd.conf.
> 
> Actually what interests me more about this benchmark (which as you say -
> are starting to get meaningless now anyhow) is the XML configuration. Would
> people prefer to setup mod_perl using some sort of XML configuration,
> because I might be interested in doing this, if there's interest.
> 
> Also, what's different between Resin's smart caching and mod_perl's? Is it
> just like StatINC?
> 
> -- 
> 
> 
> Details: FastNet Software Ltd - XML, Perl, Databases.
> Tagline: High Performance Web Solutions
> Web Sites: http://come.to/fastnet http://sergeant.org
> Available for Consultancy, Contracts and Training.

Re: Caucho faster than mod_perl?

[Perrin Harkins]

Bill Moseley wrote:
> Is Apache::Registry really that much of a hit in performance over a plain
> old Apache content handler?  And if so, why?

No, it isn't.  This is a "Hello World" benchmark we're talking about
here, designed to show the difference in startup costs.  A real app
would not have such a big difference in performance, although there
would be some difference.

Regardless, I encourage people to code for the Apache API rather than
Apache::Registry.  Besides the performance, it's just cleaner, and it
avoids some problems that can happen when you turn your program into a
sub, like accidental closures.

- Perrin

Re: XML Configuration [Was: Re: Caucho faster than mod_perl?]

[William P. McGonigle]

--- Matt Sergeant wrote:
Would
people prefer to setup mod_perl using some sort of XML configuration,
because I might be interested in doing this, if there's interest.
--- end of quote ---

I can't say it'd add much to doing it by hand, but it'd probably make writing 
configuration assistants a whole lot easier, which would be a boon for mod_perl (I've 
setup mod_perl for so many otherwise-smart friends who just couldn't get it going).

-Bill

Re: Caucho faster than mod_perl?

[Bill Moseley]

At 05:42 PM 02/01/00 -0800, Perrin Harkins wrote:
>It is not faster than mod_perl.  We tested it and found that the Resin
>author used code that didn't take good advantage of mod_perl, i.e. 
>Apache::Registry scripts rather than PerlHandler modules.

I looked at Apache::Registry and it doesn't seem like there is that much
going on per request besides a stat and an eval, and those things seem
insignificant compared to a reasonably hard working Registry script.

Is Apache::Registry really that much of a hit in performance over a plain
old Apache content handler?  And if so, why?



Bill Moseley
mailto:[EMAIL PROTECTED]

XML Configuration [Was: Re: Caucho faster than mod_perl?]

[Matt Sergeant]

On Wed, 02 Feb 2000, Perrin Harkins wrote:
> > I'm not trying to belittle Resin in any way -- in fact I'm impressed by
> > both its design and performance.  In learning about Resin during these
> > tests, I found that JSP is in many ways easier to use than mod_perl. The
> > smart caching that Resin does (with compiling .java -> .class, etc.) is
> > effective, and the XML configuration is a joy to deal with compared to
> > Apache's httpd.conf.

Actually what interests me more about this benchmark (which as you say -
are starting to get meaningless now anyhow) is the XML configuration. Would
people prefer to setup mod_perl using some sort of XML configuration,
because I might be interested in doing this, if there's interest.

Also, what's different between Resin's smart caching and mod_perl's? Is it
just like StatINC?

-- 


Details: FastNet Software Ltd - XML, Perl, Databases.
Tagline: High Performance Web Solutions
Web Sites: http://come.to/fastnet http://sergeant.org
Available for Consultancy, Contracts and Training.

Re: Caucho faster than mod_perl?

[Perrin Harkins]

Sean Chittenden wrote:
> 
> 
> Supposedly, according to its benchmarks, it's faster than mod_perl...
> impressive to say the least.  Any chance someone has any experience with this
> or would like to benchmark this technology?  External validation would be
> pretty useful.  --SC

It is not faster than mod_perl.  We tested it and found that the Resin
author used code that didn't take good advantage of mod_perl, i.e.
Apache::Registry scripts rather than PerlHandler modules.  We sent our
results and code to Caucho, and received a cordial reply, but they have
not updated their page.  The following is an excerpt from the message
one of our engineers sent to them:

>  I've recently been testing Resin 1.1b4's performance compared with
> mod_perl. The "benchmark.jar" from your web site included perl scripts that
> really didn't take full advantage of the capabilities that mod_perl offers.
> So, I wrote three perl modules that attempt to be similar to examples
> included in the Resin 1.1b4 distribution and better reflect what mod_perl
> can do.
> 
>  Attached is a gzipped tar file with these perl modules, and you're free to
> do whatever you want with them.
> 
>  Overall, I've found Resin to be a very good product, yet ran into a few
> problems during testing. These may be caused by the IBM 1.1.8 JDK rather
> than the Resin package itself. For example when running Resin as a
> stand-alone httpd server, then using "httperf" to benchmark it, the entire
> JVM would freeze up when the --close-with-reset option was given to
> httperf.
> 
>  The JVM used in testing is IBM 1.1.8, on Red Hat Linux 6.0, 2x500MHz
> Pentium III, 512MB RAM, Oracle 8.0.5 client libraries (JDBC).  Apache 1.3.9
> with DSO support was used for "Resin+Apache", with 60 max clients. A recent
> version of mod_perl and Apache  (without DSO) was used for the
> "mod_perl+Apache" tests, also with 60 max clients. Finally, "Resin httpd"
> is Resin running its own web server, with max threads set to 60.
> 
>  These tests used httperf with persistent connections, making either 1000
> or 100 requests per connection, depending on the test. The results on
> "hello" are astounding:
> 
> Test: "hello" -- displays "Hello, World"
> Size: 450 bytes
>  httperf: 40 concurrent connections, 1000 requests/connection
>  Results:
>   mod_perl+Apache: 1600 req/sec
>   Resin+Apache: 500 req/sec
>   Resin httpd: 4600 req/sec
> 
> Test: "bighello" -- lots of HTML + "Hello World"
> Size: 23888 bytes
>  httperf: 40 concurrent connections, 100 requests/connection
>  Results:
>   mod_perl+Apache: 480 req/sec
>   Resin+Apache:300 req/sec
>   Resin httpd: 280 req/sec
> 
> Test: "database" -- query: "select NAME from DIVISION", returns 11 rows
> Size: 460 bytes
>  httperf: 40 concurrent connections, 100 requests/connection
>  Results:
>   mod_perl+Apache: 570 req/sec
>   Resin+Apache:300 req/sec
>   Resin httpd: 450 req/sec
> 
> So, although "Resin httpd" performs very well with many small transfers
> ("hello"), the performance relative to mod_perl appears to degrade with
> larger transfers ("bighello"). The database performance improved nicely
> from where it was under Resin+Apache, but mod_perl still has the edge
> there.
> 
> I also tested the "file.jsp", which I renamed "fortune" in the perl tests.
> With http_load, the results again show mod_perl ahead:
> 
> Resin:
>1584 fetches, 10 max parallel, 489610 bytes, in 10 seconds
>309.097 mean bytes/connection
>158.4 fetches/sec, 48961 bytes/sec
> 
> mod_perl:
>6190 fetches, 10 max parallel, 1.98814e+06 bytes, in 10 seconds
>321.186 mean bytes/connection
>619 fetches/sec, 198814 bytes/sec
> 
> I'm not trying to belittle Resin in any way -- in fact I'm impressed by
> both its design and performance.  In learning about Resin during these
> tests, I found that JSP is in many ways easier to use than mod_perl. The
> smart caching that Resin does (with compiling .java -> .class, etc.) is
> effective, and the XML configuration is a joy to deal with compared to
> Apache's httpd.conf.

So, it's nice, but not as fast as mod_perl on Linux.  If anyone wants to
try this on another OS, I can supply a tar of the files we used for
testing.

- Perrin

Re: Caucho faster than mod_perl?

[Ask Bjoern Hansen]

On Tue, 1 Feb 2000, Sean Chittenden wrote:

>   Hey.  This is kind of relevant regarding the latest hello benchmarks
> that were released.  I was sent this today and thought it would be of some
> interest to you guys, or at least those interested in benchmarks.
> 
>   

the "loop" test indicates that they're not really knowing what they're
doing with the mod_perl side of things.

a loop should not get slower like that, so it's probably because they have
some thing that prints (which turns into r->print if I remember correctly)
a zillion lines. 

(other peoples) benchmarks sucks.  Use the tool that serves you best.


 - ask

-- 
ask bjoern hansen - 
more than 60M impressions per day, 

Re: Caucho faster than mod_perl?

[Perrin Harkins]

Sean Chittenden wrote:  
> 
> 
> Supposedly, according to its benchmarks, it's faster than mod_perl... 
> impressive to say the least.  Any chance someone has any experience
> with this or would like to benchmark this technology?  External validation
> would be pretty useful.

It is not faster than mod_perl.  We tested it and found that the Resin
author used code that didn't take good advantage of mod_perl, i.e. 
Apache::Registry scripts rather than PerlHandler modules.  We sent our
results and code to Caucho, and received a cordial reply, but they have
not updated their page.  The following is an excerpt from the message one
of our engineers sent to them: 

>  I've recently been testing Resin 1.1b4's performance compared with
> mod_perl. The "benchmark.jar" from your web site included perl scripts
> that really didn't take full advantage of the capabilities that mod_perl
> offers. So, I wrote three perl modules that attempt to be similar to
> examples included in the Resin 1.1b4 distribution and better reflect
> what mod_perl can do. 
> 
>  Attached is a gzipped tar file with these perl modules, and you're free
> to do whatever you want with them. 
> 
>  Overall, I've found Resin to be a very good product, yet ran into a few
> problems during testing. These may be caused by the IBM 1.1.8 JDK rather
> than the Resin package itself. For example when running Resin as a
> stand-alone httpd server, then using "httperf" to benchmark it, the
> entire JVM would freeze up when the --close-with-reset option was given
> to httperf. 
> 
>  The JVM used in testing is IBM 1.1.8, on Red Hat Linux 6.0, 2x500MHz
> Pentium III, 512MB RAM, Oracle 8.0.5 client libraries (JDBC).  Apache
> 1.3.9 with DSO support was used for "Resin+Apache", with 60 max clients. 
> A recent version of mod_perl and Apache (without DSO) was used for the
> "mod_perl+Apache" tests, also with 60 max clients. Finally, "Resin
> httpd" is Resin running its own web server, with max threads set to 60. 
> 
>  These tests used httperf with persistent connections, making either
> 1000 or 100 requests per connection, depending on the test. The results
> on "hello" are astounding: 
> 
> Test: "hello" -- displays "Hello, World" 
> Size: 450 bytes
>  httperf: 40 concurrent connections, 1000 requests/connection
>  Results: 
>   mod_perl+Apache: 1600 req/sec
>   Resin+Apache:  500 req/sec
>   Resin httpd:  4600 req/sec
> 
> Test: "bighello" -- lots of HTML + "Hello World" 
> Size: 23888 bytes
>  httperf: 40 concurrent connections, 100 requests/connection
>  Results: 
>   mod_perl+Apache: 480 req/sec
>   Resin+Apache:  300 req/sec
>   Resin httpd:  280 req/sec
> 
> Test: "database" -- query: "select NAME from DIVISION", returns 11
> rows
> Size: 460 bytes
>  httperf: 40 concurrent connections, 100 requests/connection
>  Results: 
>   mod_perl+Apache: 570 req/sec
>   Resin+Apache:  300 req/sec
>   Resin httpd:  450 req/sec
> 
> So, although "Resin httpd" performs very well with many small transfers
> ("hello"), the performance relative to mod_perl appears to degrade with
> larger transfers ("bighello"). The database performance improved nicely
> from where it was under Resin+Apache, but mod_perl still has the edge
> there. 
> 
> I also tested the "file.jsp", which I renamed "fortune" in the perl
> tests.  With http_load, the results again show mod_perl ahead: 
> 
> Resin: 
>1584 fetches, 10 max parallel, 489610 bytes, in 10 seconds
>309.097 mean bytes/connection
>158.4 fetches/sec, 48961 bytes/sec
> 
> mod_perl: 
>6190 fetches, 10 max parallel, 1.98814e+06 bytes, in 10 seconds
>321.186 mean bytes/connection
>619 fetches/sec, 198814 bytes/sec
> 
> I'm not trying to belittle Resin in any way -- in fact I'm impressed by
> both its design and performance.  In learning about Resin during these
> tests, I found that JSP is in many ways easier to use than mod_perl. The
> smart caching that Resin does (with compiling .java -> .class, etc.) is
> effective, and the XML configuration is a joy to deal with compared to
> Apache's httpd.conf.

So, it's nice, but not as fast as mod_perl on Linux.  If anyone wants to
try this on another OS, I can supply a tar of the files we used for
testing. 
 - Perrin

Re: Caucho faster than mod_perl?

[Joshua Chamas]

Sean Chittenden wrote:
> 
> Hey.  This is kind of relevant regarding the latest hello benchmarks
> that were released.  I was sent this today and thought it would be of some
> interest to you guys, or at least those interested in benchmarks.
> 
> 
> 
> Supposedly, according to its benchmarks, it's faster than mod_perl...
> impressive to say the least.  Any chance someone has any experience with this
> or would like to benchmark this technology?  External validation would be
> pretty useful.  --SC
> 
> --
> Sean Chittenden

I'll be benchmarking it on WinNT, which should give a good
relative value to the mod_perl environments I've tested.
Note that in the "hello world" benchmarks at 

  http://www.chamas.com/bench/hello_bysystem.html

Scott from Caucho only benchmarked Registry scripts, but
didn't do the native mod_perl handler test, which runs
significantly faster, anywhere from 50% to 100% faster on
Linux from what I can tell, which would put it right next
to Resin's JSP Java for HelloWorld.

It would be great if someone with a Linux box could do
both a modperl handler & a resin jsp.  If Chip, Kimbro, or Dirk
submitted the bench, that would be great!, since they already
done a few on their Linux boxes to readily compare against.

-- Joshua
_
Joshua Chamas   Chamas Enterprises Inc.
NodeWorks >> free web link monitoring   Huntington Beach, CA  USA 
http://www.nodeworks.com1-714-625-4051

Caucho faster than mod_perl?

[Sean Chittenden]

Hey.  This is kind of relevant regarding the latest hello benchmarks
that were released.  I was sent this today and thought it would be of some
interest to you guys, or at least those interested in benchmarks.



Supposedly, according to its benchmarks, it's faster than mod_perl... 
impressive to say the least.  Any chance someone has any experience with this
or would like to benchmark this technology?  External validation would be
pretty useful.  --SC

-- 
Sean Chittenden