
The Beowulf Scalable Mass Storage Project will assemble, using all commodity off-the-shelf (COTS) components, a trivially reproducible system that provides a terabyte of secondary storage and a gigabyte/s aggregate external bandwidth. The system software will use enhancements to the Linux operating system to integrate and extend public available parallel I/O and file systems. The system will be evaluated on real applications in the operational environment provided by the Earth and Space Sciences (ESS) Project within NASA's HPCC Program.
The Beowulf Scalable Secondary Storage project will build on the technology and methodology originally developed as part of the Beowulf Parallel Workstation project. The availability of the Linux operating system and the performance levels attend by PC-market components has created the opportunity to provide a cost-effective solutions to secondary storage problem.
Linux is a publicly available POSIX compliant operating system developed for the x86 architecture. Linux allows us to easily harvest the software built for clusters of workstation, and since it is distributed with source, we can enhance the systems to optimize its performance. The Beowulf development team is an active member of the Linux develop community and a member of the growing community of research teams that are finding solutions to the computing demands using the pile-of-PC's approach.
The system will be developed within the infrastructure established to support the Grand Challenge Investigators for the second phase of the ESS Project. This will allow us to the test in an operational environment. The Beowulf project has been identified by the Scalable I/O Initiative as a target testbed. This direct involvement with a large number of diverse institutions either requiring or developing high-performance I/O systems will provide independent performance evaluation and be a vehicle for technology transfer. The primary vehicle will be Linux itself; any enhancements to the operating system or application software will be made available on the Web through the usual Linux distribution mechanisms.

This is a 3-year project that started at the end of 96. We have initiated testing of key hardware and software components.
Dr. Phillip R. Merkey
Center of Excellence in Space Data and Information Sciences
(CESDIS)
Goddard Space Flight Center
merk@cesdis.gsfc.nasa.gov
301-286-3805
Table of Contents | Subsection Contents -- Testbeds