Network Storage Virtualization for Data Intensive Collaboration
Time 04/25/07 08:45AM-10:00AM
The data intensive nature of today's research and multimedia projects substantially increases the need to integrate networking and storage technologies in a flexible, high performance way. Such an improved synthesis of networking and storage is now being deployed by the Research and Education Data Depot Network (REDDnet) project, an NSF funded infrastructure project that aims to support data intensive collaboration among participating research and education application communities. Over the next year and a half, REDDnet will provision roughly 700TB of distributed storage across several sites on US research networks. Using Logistical Networking (LN) middleware, which builds on the Internet Backplane Protocol (IBP) for storage virtualization, REDDnet enables users to exploit a high degree of parallelism in network transfers and disk I/0, while at the same time providing for easy deployment, flexible application development, and a sustainable approach to interoperability. After a brief overview of REDDnet and its key technologies, this session will feature short talks by REDDnet application community leaders, who will describe and demonstrate how this national cyberinfrastructure can be used to support distributed data management and collaborative applications. Application areas represented will include Particle Physics, Remote Earth Sensing, and dynamic sharing of large scale visualization.
Speaker Paul Sheldon Vanderbilt University
Speaker Micah Beck University of Tennessee
Speaker Paul R. Blackwell Stephen F Austin State University
Speaker Jian Huang University of Tennessee
Speaker Alan Tackett Vanderbilt University
Secondary tracks Advancing Applications: What's Working? What's Not? What's Needed?