So... I decided to use JB 6 for a new project I'm working on, and decided to move over some VERY STABLE code that has been under Spring IoC control. All I can say is 24+ hours of wasted time later, I decided to give it a shot in GlassFish. It worked like a charm...
It seems that JB6 has some strange assumptions about activating singletons and EJB injections that do not make sense.... and it was starting up things before dependencies were deployed and then complaining about JNDI resolution failures.
Later on... in the server logs, you can see that it deploys the EJB's in question. So... I wonder why it couldn't identify the dependencies and deploy them as needed... since GlassFish is able to do it...
Strange...
It seems that JB6 has some strange assumptions about activating singletons and EJB injections that do not make sense.... and it was starting up things before dependencies were deployed and then complaining about JNDI resolution failures.
Later on... in the server logs, you can see that it deploys the EJB's in question. So... I wonder why it couldn't identify the dependencies and deploy them as needed... since GlassFish is able to do it...
Strange...
