Multicasting is an effective method to guarantee scalability of data transfer. Multicast applications range from the relief of Internet hot spots to healthcare alert systems. Much research has focused on isolated data management issues that arise in a multicast environment, including our previous work on caching, scheduling, indexing, hybrid schemes, and consistency maintenance. This paper discusses the integration of these research contributions and the transition to a working software distribution that provides the middleware support of a data management layer to applications. Our middleware is flexible, can be shared across applications, and operates on top of existing and upcoming implementations of multicast protocols. The middleware benefits distributed applications with a uniform, efficient, scalable, and state-of-the-art support for critical data management functionality.
Panos K. Chrysanthis, Vincenzo Liberatore, Kirk Pr