The logs shows how it's supposed to be. :-)
I'm missing the AppClientParsingDeployer. My guess is it will be run
before the AppClientScanningDeployer (as it is now, but not how it
ultimately should be).
The icky part is the EJBRegistrationDeployer telling the
SecurityDeployer to shut up (EJBRegistrationDeployer@line 167). This is
because we don't have a EjbScanningDeployer yet, so the SecurityDeployer
ignores annotations. Somewhere in the EJB 3 code it creates it's own
Jacc policy and stuff. For now this would mean the SecurityDeployer
should go after EJBRegistrationDeployer for icky reasons.
The output of EJBRegistrationDeployer is Ejb3Deployment +
"org.jboss.deployment.security.ignoreMe" and the input of
SecurityDeployer should be: WebMetaData | ApplicationMetaData |
BeanMetaData + "org.jboss.deployment.security.ignoreMe".
Carlo
On Fri, 2007-06-29 at 13:06 +0200, Adrian wrote:
Carlo, can you validate that the deployment ordering
shown in the DEBUG message here is correct for EJB3
http://www.jboss.com/index.html?module=bb&op=viewtopic&t=112181&a...
I sort of guessed it from what attachments the deployers were
getting/adding. :-)
On Fri, 2007-06-29 at 12:21 +0200, Adrian wrote:
> On Fri, 2007-06-29 at 11:14 +0200, Carlo de Wolf wrote:
> > You're welcome. :-)
> >
> > Just out of curiosity, what did you test?
> > (
http://wolf.nerdnet.nl:8180/hudson/job/JBoss_EJB3_TRUNK/13/ )
> >
>
> Just the basic stateless and entity standup tests.
>
> > Meow,
> >
> > Carlo
> >