users@glassfish.java.net

Re: Mystery problem w/ Wicket + Glassfish v2?

From: Gabor Szokoli <szocske_at_gmail.com>
Date: Tue, 6 Nov 2007 11:30:16 +0100

Hi,

I can offer some anecdotal reference you may find relevant, we too use
wicket 1.2 with glassfish 2 and ported from jboss 4:

On Nov 5, 2007 4:42 PM, <glassfish_at_javadesktop.org> wrote:
> Occasionally, my application will "just start doing this" - throwing this error. I am using Wicket 1.2.4 because my application is "stuck" at that version for the time being. I do not have the time to make the changes to upgrade to the latest version, right now.

Upgrading form wicket 1.2.4 to 1.2.6 was only a change in the pom.xml
to us, no need to touch the code.
We never experienced anything like you do.
We had strange persistence.xml issues (completely unlike yours) with
glassfish "v2-58c" variant, which vanished when we upgraded to "g".

> this NEVER happened on JBoss 4.0.x with Wicket. I figured it was safe to assume this had nothing to do with my code -

That's what we thought too when we moved from JBoss to GlassFish:
(obviously not the same as your problem) The subcomponents of our EAR
file were all packaged as if they were plain jars. (our mistake.)
JBoss silently deduced the intended roles from the file types and
contents. Glassfish does not. We debugged this for days.

> nowhere in the stack does it point to anything in my application (or any application running on it, for that matter.)

We don't see the whole stack trace, but it does refer to wicket
classes (which are "yours" from the application servers point of
view).


Gabor Szokoli