Distributed caching and replication

The number of people and services daily relying upon distributed information, such as in the Word-Wide Web (WWW) or distributed shared memory systems is tremendously growing. This trend will surely continue, so that the resources of those systems will be exhausted not too far from today. There are two alternatives for coping with the tremendously increasing workload of distributed information systems. One long-term strategy consists in enhancing the resources, e.g., by replacing low-bandwidth communication links by high-bandwidth ones. Unfortunately, this approach is very expensive and its realization will take many years. For some distributed information systems, like the WWW, putting this strategy to work - although unavoidable on the long run - is further complicated by the absence of a central authority being responsible for the system administration. A more promising strategy with respect to cost and time constraints is to manage available resources in a more appropriate fashion. It is the latter short-term strategy on which we focus our attention in the scope of this mini-track.