Replicated data in a distributed environment

Replication Server, a forthcoming product that dynamically maintains subsets of data in a distributed environment, providing several transaction models to maintain loose consistency, is contrasted with existing products which provide location-transparent reads and high-consistency coordinated commits. Replication Server makes it possible to build systems that are much more robust in the face of system component failures. By moving transactions rather than data, and by locating data at the point of processing, it maximizes the use of network bandwidth, enabling the deployment of robust, high-performance applications at a lower cost than with traditional approaches