"File systems and supercomputers are evolving separately, and integrating multiple generations of hardware and software is a requirement for the NAS facility. We have moved to a data-centric model where the file system is a standalone component that persists between generations of supercomputers. Lustre is the key element to make this work", stated Bob Ciotti, supercomputing systems group lead and lead system architect for the NAS Division. "Our plan with this effort will allow for better operational continuity as we continue to evolve our super-scale computational environments."
"Being chosen by NASA to support their Lustre operations is an extreme honour. NASA is a world leader in a wide range of research areas involving heavy computation. Fast and reliable data movement & storage is required to make use of the large computers", stated Brent Gorda, CEO of Whamcloud. "NASA depends on computational results for scientific discovery, design and mission operation. We're excited to be a part of supporting their research and the smooth operation of the data path."
Pleiades, NASA's largest system, ranks 7th on the TOP500 list of the world's most powerful supercomputers and has Lustre file systems that provide nearly 10 PB of storage and serve over 110 thousand cores. In a data-centric model where compute is attached to storage, Lustre provides shared data access between all major systems at the center including the hyperwall-2 visualization system and the Columbia supercomputer. Scientists are running large-scale jobs on these systems to gain insight into Earth's ocean and climate variability; reduce harmful emissions from aircraft; and design future vehicles for planetary and space exploration. The systems were also critical to supporting debris damage assessment on space shuttle missions and gave managers data about critical decisions to perform repairs and clear the orbiter for safe landing.