Design of a middleware service for scalable wide area network applications

Most distributed applications would greatly benefit from a middleware set of services to provide scalability, fault-tolerance, adaptability to network events, load balancing and real-time guarantees. However current distributed paradigms, like CORBA or RPC only provide the basic protocols and services for interoperability, and fail (intentionally) to define more complex relationships among an application´s distributed elements. We present the design of a middleware service layer that provides applications with shared common facilities to achieve the desired scalability, fault-tolerance, adaptability and load balancing properties, that otherwise need to be defined on a per application basis. Our design is not tied to any particular distributed paradigm (CORBA, RPC, etc), but allows transparent interoperability between applications originally written for them, minimizing the performance penalty incurred upon the application by our framework