Motivation

Any casual user of the Web knows that there is a critical need for scalable data transfer methods. Web servers and networks can be brought to their knees by hot-spot access patterns. Examples of Internet hot spots include news sites such as CNN after the 2000 presidential elections, the NASA site after the Mars Pathfinder landing, and the host site at the Nagano Olympics. In each case millions of clients attempted to simultaneously access the sites. Servers and networks were overwhelmed by the large number of requests, resulting in long delays to retrieve documents. One solution to this scalability problem is multicast. In these hot spot examples, many clients were interested in nearly the same data. Thus, a single server could have served all such requests by multicasting the hot data items to interested clients without any explicit request from the clients.

Our goal is to integrate data dissemination and multicast communication techniques into a working software distribution that provides the middleware support of a scalable multicast-based data management layer to applications. The time is ripe to exploit new and effective Internet multicast solutions to relieve the scalability problems of data-intensive distributed applications. Our approach will also impact data dissemination in wired and wireless local area networks, as well as satellite links, where broadcasting is the principal mode of communication.

Figure 1: The middleware data dissemination architecture, and its relationship with the application and transport layers.

[width=3.3in,height=4.0in,angle=270]newarch.eps