You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomcat.apache.org by Mladen Turk <mt...@apache.org> on 2005/04/21 15:20:22 UTC
Tomcat/APR benchmark results
Hi,
Here are the brief results for Tomcat HEAD:
Server Threads Pause (ms) Error(%) Rate (req/sec)
Apache2.0.49 500 1000 1.74 124.2
Http11Protocol 500 1000 0.20 139.5
Http11AprProtocol 500 1000 0.00 266.9
Tests has been run on SLES9/64-Bit/Java-1.5.0_02-b09
Client was WINXP/SP2 running JMeter and Peter Lin's test plans.
Those tests simulate 500 concurrent users connecting to a
150 maxClient server. The connection is keep-alive with
1 second delay between the requests.
So our JNI/APR http connectors is almost twice faster then
both Tomcat standard http connector and Apache2/prefork.
What is more important is that the error number is zero,
meaning there were no refused connections, because of
keepalive connection poller.
Regards,
Mladen.
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Peter Lin <wo...@gmail.com>.
I'll be using my AMD 2ghz linux box running Fedora Core1. haven't
updated to FC3 yet, though remy keeps suggesting I upgrade :)
peter
On 4/21/05, Mladen Turk <mt...@apache.org> wrote:
> Peter Lin wrote:
> > Cool. I assume you were finally able to my test plan mladen?
>
> Yes. I've run them extensively, thanks for sharing those.
> Been playing with the constant timer, number of threads, etc...
> Results show that for high number of connections the APR
> implementation is way ahead of any other.
> Me very happy :)
>
>
> > I plan
> > to run the benchmarks on my system this weekend, not that my laptop is
> > healthy.
> >
>
> Sure.
> If you are using WIN32 as server, you will need a patched APR
> to be able to handle more then a 64 FD_SETSIZE limit.
> Tell me, and I can mail it to you.
>
>
> Regards,
> Mladen.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Mladen Turk <mt...@apache.org>.
Peter Lin wrote:
> Cool. I assume you were finally able to my test plan mladen?
Yes. I've run them extensively, thanks for sharing those.
Been playing with the constant timer, number of threads, etc...
Results show that for high number of connections the APR
implementation is way ahead of any other.
Me very happy :)
> I plan
> to run the benchmarks on my system this weekend, not that my laptop is
> healthy.
>
Sure.
If you are using WIN32 as server, you will need a patched APR
to be able to handle more then a 64 FD_SETSIZE limit.
Tell me, and I can mail it to you.
Regards,
Mladen.
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Peter Lin <wo...@gmail.com>.
Cool. I assume you were finally able to my test plan mladen? I plan
to run the benchmarks on my system this weekend, not that my laptop is
healthy.
peter
On 4/21/05, Mladen Turk <mt...@apache.org> wrote:
> Hi,
>
> Here are the brief results for Tomcat HEAD:
>
> Server Threads Pause (ms) Error(%) Rate (req/sec)
> Apache2.0.49 500 1000 1.74 124.2
> Http11Protocol 500 1000 0.20 139.5
> Http11AprProtocol 500 1000 0.00 266.9
>
> Tests has been run on SLES9/64-Bit/Java-1.5.0_02-b09
> Client was WINXP/SP2 running JMeter and Peter Lin's test plans.
>
> Those tests simulate 500 concurrent users connecting to a
> 150 maxClient server. The connection is keep-alive with
> 1 second delay between the requests.
>
> So our JNI/APR http connectors is almost twice faster then
> both Tomcat standard http connector and Apache2/prefork.
>
> What is more important is that the error number is zero,
> meaning there were no refused connections, because of
> keepalive connection poller.
>
> Regards,
> Mladen.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Remy Maucherat <re...@apache.org>.
Mladen Turk wrote:
> Hi,
>
> Here are the brief results for Tomcat HEAD:
>
> Server Threads Pause (ms) Error(%) Rate (req/sec)
> Apache2.0.49 500 1000 1.74 124.2
> Http11Protocol 500 1000 0.20 139.5
> Http11AprProtocol 500 1000 0.00 266.9
>
> Tests has been run on SLES9/64-Bit/Java-1.5.0_02-b09
> Client was WINXP/SP2 running JMeter and Peter Lin's test plans.
>
> Those tests simulate 500 concurrent users connecting to a
> 150 maxClient server. The connection is keep-alive with
> 1 second delay between the requests.
Can you also test after increasing maxThreads to something really high
(and increase the number of concurrent users as well) ? This should
improve performance for Http11Protocol. Then you can introduce other
metrics, like memory usage or whatever :)
Rémy
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Mladen Turk <mt...@apache.org>.
Jess Holle wrote:
> Would we expect anything markedly different with Apache 2 worker rather
> than prefork?
No.
Also I found that the majority of users is usig prefork because of PHP.
> This would seem a bit more of an apples-to-apples
> comparison (i.e. comparing multi-threaded low-process count servers).
> Also how many simultaneous workers (threads/processes/whatever) were
> allowewd in each case?
>
150 maxThreads/MaxClients in all cases.
The purpose of this connector is to deal with keep-alive connections,
so that's why the test were run with 500
cuncurrent connections with 1 second delay between each request/thread.
Of course we can set the MaxClients to 500, but what to do with
5000 or more of them?
For a smaller number of threads (50) using ab without a delay between
keepalive requests:
Server Req/sec
HTTPD/2.0.49 4740.74
Http11Protocol 5203.25
Http11AprProtocol 5203.25
Request is for the tomcat.gif (1943 bytes file).
Regards,
Mladen.
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Peter Lin <wo...@gmail.com>.
I'll try running the same jmeter test plan on apache2 this weekend,
but I'm guessing the difference will be minimal. based on the results
for the static file benchmarks, the delta between 1.3.x and 2.x wasn't
significant. I think those numbers are in the excel spreadsheet I
posted
http://people.apache.org/~woolfel/tc_results.html
if the results aren't in the spreadsheet I'll try to dig up the
numbers for 1.3. I could be wrong, but the difference between httpd
1.3 and 2.0 was about 30-35%
peter
On 4/21/05, Jess Holle <je...@ptc.com> wrote:
> Would we expect anything markedly different with Apache 2 worker rather
> than prefork? This would seem a bit more of an apples-to-apples
> comparison (i.e. comparing multi-threaded low-process count servers).
> Also how many simultaneous workers (threads/processes/whatever) were
> allowewd in each case?
>
> [I stopped using prefork on all platforms when moving to Apache 2.]
>
> --
> Jess Holle
>
> Mladen Turk wrote:
>
> > Hi,
> >
> > Here are the brief results for Tomcat HEAD:
> >
> > Server Threads Pause (ms) Error(%) Rate (req/sec)
> > Apache2.0.49 500 1000 1.74 124.2
> > Http11Protocol 500 1000 0.20 139.5
> > Http11AprProtocol 500 1000 0.00 266.9
> >
> > Tests has been run on SLES9/64-Bit/Java-1.5.0_02-b09
> > Client was WINXP/SP2 running JMeter and Peter Lin's test plans.
> >
> > Those tests simulate 500 concurrent users connecting to a
> > 150 maxClient server. The connection is keep-alive with
> > 1 second delay between the requests.
> >
> > So our JNI/APR http connectors is almost twice faster then
> > both Tomcat standard http connector and Apache2/prefork.
> >
> > What is more important is that the error number is zero,
> > meaning there were no refused connections, because of
> > keepalive connection poller.
> >
> >
> > Regards,
> > Mladen.
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
> > For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
Re: Tomcat/APR benchmark results
Posted by Jess Holle <je...@ptc.com>.
Would we expect anything markedly different with Apache 2 worker rather
than prefork? This would seem a bit more of an apples-to-apples
comparison (i.e. comparing multi-threaded low-process count servers).
Also how many simultaneous workers (threads/processes/whatever) were
allowewd in each case?
[I stopped using prefork on all platforms when moving to Apache 2.]
--
Jess Holle
Mladen Turk wrote:
> Hi,
>
> Here are the brief results for Tomcat HEAD:
>
> Server Threads Pause (ms) Error(%) Rate (req/sec)
> Apache2.0.49 500 1000 1.74 124.2
> Http11Protocol 500 1000 0.20 139.5
> Http11AprProtocol 500 1000 0.00 266.9
>
> Tests has been run on SLES9/64-Bit/Java-1.5.0_02-b09
> Client was WINXP/SP2 running JMeter and Peter Lin's test plans.
>
> Those tests simulate 500 concurrent users connecting to a
> 150 maxClient server. The connection is keep-alive with
> 1 second delay between the requests.
>
> So our JNI/APR http connectors is almost twice faster then
> both Tomcat standard http connector and Apache2/prefork.
>
> What is more important is that the error number is zero,
> meaning there were no refused connections, because of
> keepalive connection poller.
>
>
> Regards,
> Mladen.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org