I want to develop a calculation engine to perform various calculations,
using the J2EE framework. Although at first sight, it seems like it does
not make sense to have more than one calculation engine, scalability may
be an issue so I would like to have a pool of these engines that deal
with calculation requests from remote clients (client side). However, it
would obviously be overkill to have a seperate instance of the
calculation engine for each client that uses the calculation.

This seems (in my mind) to fall into a commonly occuring problem domain
in distributed computing. How can I use EJBs to work out a solution? (a
pool of stateless calculation engines handling calculation requests from
remote clients).

Any pointers to resources/code will be very useful. As an aside (i.e. a
further question), when implementing the methods of the calculation
engine class, should the methods be declared as static (i.e. class methods?)

Many thanks