Hi Brice, That interface is already there and is called WorkerPool.
Frank Brice Fines wrote: > Hi Frank, > > You are right, double pooling would not make sense (having ThreadPool > getting threads from the server was not a good idea), but it is not my > aim. > My main concern is not to start application-threads in a J2EE environment. > Starting applications threads is unsupported and sometimes not allowed by > J2EE servers. There seems to be various reason for that: the container is > unaware of application-threads and will not be able to control them/shut > them down, transaction support, cluster support, resource allocation > management on server, scalability,...). According to EJB 2.1 > specifications it is not allowed, according to servlet 2.3/2.4 it is > unsupported and not recommended. > > Thus, maybe ThreadPool could become an interface, and, this way, we could > have an implementation using server resources (a J2EE compatible > ThreadPoolImpl) and pass it to the MultiThreadedMessageDispatcher > constructor. > Do you thing that could be a solution? > > I will have a look a SNM4J-Agent. > Thanks > > Regards > Brice > > (some links on the application threads in J2EE topic: > http://www.theserverside.com/discussions/thread.tss?thread_id=44353 > http://stackoverflow.com/questions/533783/why-spawning-threads-in-j2ee-container-is-discouraged > > http://www.onjava.com/pub/a/onjava/2004/10/13/j2ee-timers.html?page=1 > http://e-docs.bea.com/wls/docs100/programming/topics.html > ) > _______________________________________________ > SNMP4J mailing list > [email protected] > http://lists.agentpp.org/mailman/listinfo/snmp4j -- AGENT++ http://www.agentpp.com http://www.snmp4j.com http://www.mibexplorer.com http://www.mibdesigner.com _______________________________________________ SNMP4J mailing list [email protected] http://lists.agentpp.org/mailman/listinfo/snmp4j
