Usecase to EJB Component Mapping

[Pearlo Muthukumaran]

Hi,
I have had a good deal of search on web to find out how the Usecases get Mapped onto Beans. Some article at www.theserverside.com gives a very vague guideline that in a usecase sequence diagram, the usecase itself maps to a Session Bean (SLSB, SFSB) and a entities that exchange messages in digram maps onto either entity or session bean once again(such session bean will turn out to be a <<include>> type of use cases which is to be exploded further to fine-grain the domain model - which is understandable). My questions are
1. Is the above a generic right approach?
2. Is there any special requirement on the usecase design itself in order to accomodate above way of mapping? and
3. Is there any other approach for carrying out this mapping?
Thanks in advance
Muthu

[Kyle Brown]

I totally disagree that each use case becomes a session bean. That results in wayyyy too many session beans. See the advice at the bottom of this article for my recommendations.
Kyle