No gain after pre-compilation

2010-03-09 Thread Alexander Skwar
Hello.

In the Tomcat 6.0 Jasper documentation at
http://tomcat.apache.org/tomcat-6.0-doc/jasper-howto.html#Production%20Configuration
or http://LNK.by/c2H2, I read that the most performance gain can be
achieved by pre-compiling the JSPs. Because of that, I simply gave that
a try and can't find much of a gain at all.

I used Ant 1.8.0 with Java 1.6.0_18 and the build.xml script at
http://pastebin.ca/1828486
We're doing around the clock performance measurements, where some
Watir based script runs a browser and calls pre-defined pages and tasks
in the web app. We measure, how long this takes. Please see
http://goo.gl/cokf for a graph showing this. Yesterday, at about 16:00,
I installed the pre-compiled web app. There's no (visible) change in the
measurements to be found.

Did I do something wrong? Was the build.xml not good (enough)? I invoked
it with:

   tomcat_home=/usr/share/tomcat
   webapp_path=/var/webapp
   ant -Dtomcat.home=$tomcat_home -Dwebapp.path=$webapp_path

Or is it just, that this shows, that our web app doesn't have a bottleneck
in the JSP land (which would be a true statement - it communicates
with some background databases and apps, which might not be the
fastest...)?

One thing which confuses me a bit though, is that when I call the web
app for the first time (after deployment or after having re-started Tomcat),
the web app is slow. Ie. I go to http://server:8080/webapp and wait for
like 10 seconds until the login screen of the web app is shown. This
wait only happens for the very first time after deployment. When I do
this again a few minutes/hours later, http://server:8080/webapp is shown
right away.

Shouldn't pre-compilation also make this first time wait go away?

Thanks a lot,

Alexander

-
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org



Re: No gain after pre-compilation

2010-03-09 Thread Alexander Skwar

Hello!

Thanks a lot for the conformation.

Do you happen to have any idea, why I did not notice any impovements  
in the 1st time load delay? Ie. why do I still have to wait 10s after  
having deployed, before the login screen is shown for the first time?


At this point in time, the web app does not require the background  
daemons at all. The login page is even shown, when those daemons are  
down.


Best regards,

Alexander

Am 09.03.2010 um 10:54 schrieb Markus Schönhaber tomcat-us...@list-post.mks-mail.d 
e:



09.03.2010 10:25, Alexander Skwar:

I didn't look at your build script, just a general remark below.

Or is it just, that this shows, that our web app doesn't have a  
bottleneck

in the JSP land (which would be a true statement - it communicates
with some background databases and apps, which might not be the
fastest...)?

One thing which confuses me a bit though, is that when I call the web
app for the first time (after deployment or after having re-started  
Tomcat),
the web app is slow. Ie. I go to http://server:8080/webapp and  
wait for

like 10 seconds until the login screen of the web app is shown. This
wait only happens for the very first time after deployment. When I do
this again a few minutes/hours later, http://server:8080/webapp is  
shown

right away.

Shouldn't pre-compilation also make this first time wait go away?


Not also - the *only* gain that can be expected by pre-compilation  
of

JSPs is the elimination of the first time wait, since Tomcat doesn't
need to do the compilation on-the-fly when the JSP is accessed for the
first time. For all following accesses of the JSP it shouldn't matter
whether or not it was pre-compiled.
But: a noticeable difference can only be expected if it really is the
compilation that slows things down on first access. If, OTOH, the vast
amount of time is taken by, say, creation of DB connections,
pre-compiling the JSPs won't make much of a difference.

--
Regards
 mks

-
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org



-
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org



High Load examples?

2010-01-07 Thread Alexander Skwar

Hello.

Right now, we're using Bea WebLogic as our application server. We'd like to
change to Tomcat 6. Now the creator of the application being run on the
App Server said, that Tomcat works very well under low usage / low load
situations. But if there are many (unsure about the definition of
many...)
users, Tomcat might tend to not perform so well anymore.

I suppose many might mean like 50-100+ concurrent users using the App
or system at the same time.

Now I'm wondering, if that's actually true - does Tomcat not perform well
anymore, if there are many users using it? Does anyone of you maybe have
real world examples of high profile / high load sites using Tomcat (6)?

Best regards,
Alexander
-- 
View this message in context: 
http://old.nabble.com/High-Load-examples--tp27058015p27058015.html
Sent from the Tomcat - User mailing list archive at Nabble.com.


-
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org



Incompatabilities?

2010-01-07 Thread Alexander Skwar

Good Morning!

Thanks a lot for all the answers you provided yesterday in my other thread,
asking about scalability of Tomcat. Appreciate it!

Besides the load issue, they claim, that a different app of theirs doesn't
work with Tomcat, but it's supposed to work with WebLogic and jBoss. Am
I right in assuming, that this is most probably caused by bugs in their app,
which (somehow only) get triggered by Tomcat?

Can't be any more specific here, sorry. I don't know what kind of errors
these
(supposedly...) are and I'd rather also not name the App and Manufacturer
(only
so much: it's by far not a small company making the app and the app is used
quite a lot in its field and, of course, it ain't cheap...).

I'm simply asking to figure out the validity of their claims.

Thanks a lot again,

Alexander
-- 
View this message in context: 
http://old.nabble.com/Incompatabilities--tp27072171p27072171.html
Sent from the Tomcat - User mailing list archive at Nabble.com.


-
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org