Distributed storage, or shared data, is a vital mechanism for communication among processors in distributed systems. The use of shared memory allows for better structured and easier-to-verify distributed applications. This shared service facilitates the development of higher-level applications including collaborative editing systems, multiplayer games and scientific data repositories.
Shared data is not provided off the shelf in large-scale distributed systems. Instead, processors must keep individual copies of the data. These processors communicate by sending messages to keep the replicas consistent resulting in reduced performances.
Jennifer Welch, Regents Professor and Chevron Professor II in the Department of Computer Science and Engineering at Texas A&M University, is currently exploring how relaxing the specifications of the shared data can improve the performance of distributed systems.
More at the Dwight Look College of Engineering