tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sven Köhler <>
Subject Re: AW: modified jsp-files and jasper2
Date Sat, 15 Feb 2003 13:30:55 GMT
> I just checked the web.xml file in $CATALINA_HOME/conf/web.xml.  It suggests
> the "development" is the default mode in which Jasper operates.  Have you
> tried leaving development, reloading, and checkInterval out of your web.xml
> file to see if the standard behavior works?

The tomcat-docs also state, that development should be false in 
production-environment. background should be used, because it's more 

Even with development=true i remember some cases, where tomcat didn't 
reload the pages, and i had to hit the reload-button.

>>Hitting the reload-button in the manager-app always helps.
> It shouldn't help; it either works or it doesn't.  You should also note that
> there can be several seconds lag associated with having Tomcat actually
> carry out a reload.
> [ Aside:  Does manager/reload fail if Context.reloadable=false? ]

"reloadable" should be renamed into "autoreload".
that's what it is for.
if reloadable==true then tomcat checks the files in the WEB-INF-folder 
for modifications.

> You have not mentioned your reloadable settings in server.xml.  I am
> inclined to believe that reloading class files for a web application is
> handled in one place--no matter where those file reside.

it is false, because i don't want "auto-reload".

> I suggested 60 seconds for this...
> 1) There has been a change.
> 2) Jasper translates the file.
> 3) Jasper compiles the file.
> 4) Tomcat takes up to 15 seconds at this point to determine that it needs to
> reload a class file.
> I wasn't implying that the checkInterval number was useless--just that there
> is more processing to reloading a class than mere class file datetime compar
> isons.

OK, but a few hours time should be enough.

> By your tone, it sounds like you only conducted this experiment once.

No, my client calls me all the time, and it's anoying.
It seems that i suffer a bug that nobody notices or cares about.

> If I were you, I would start with the most simple system for base testing
> and then add complexity in increments.  Consider goes with these phases:

I will do some tests when i have the time.

> Reloading of applications does work for lots of people.  I am sure that if
> you start from the baseline, you will discover something that was not
> previously understood about how the reloading process works or perhaps work
> a kink out of your configuration.

It cannot be my fault.
There has to be a bug. As i already told you, it sometimes work, and 
semetimes it doesn't.
I havan't noticed a tomcat-option which tells tomcat to randomize reloads.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message