Originally posted by Ramon Gill:
Some people have used notes in the class diagrams to say what are entity beans, session beans, etc.
Originally posted by Priya Patel:
I did not show *ANY* J2EE extension classes.
Originally posted by Ajith Kallambella:
A domain model is normally technology agnostic. Technology here includes both implementation platform such as J2EE and analysis methodology such as OO.
After the business analysts finalize the domain model, you bring in the technology experts to decompose the model by applying a specific methodology. In case of a typical OO analysis phase you do things such as identifying generalization/specializations, separating processes from participants, defining contracts( interfaces ), identify conceptual divisions in order to define types and so on.
Once this is done, you apply platform specific(J2EE) principles and identify implementation details/constraints - things such as what gets modelled as entity beans( and why ), what can be just a POJO etc.
So to answer your original question( finally ), you should ensure that, during the steps detailed above, you do not contradict any relationships specified in the original domain model. This is an absolute must. Avoid diluting the clarity of any relationship and don't redefine the cardinality.