Terascala Streamlines File System Management at Virginia Tech

Share Article

Terascala’s TeraOS Simplifies Lustre administration, saving time and allowing users to focus on research

News Image
Terascala’s TeraOS provides a truly complete Lustre parallel file system solution.-- Brian Marshall, computational scientist at Virginia Tech

Terascala, the industry leader in High Performance Computing (HPC) storage management software, today announced that a high-performance storage appliance managed by Terascala is simplifying Lustre® file system management at Virginia Tech’s Advanced Research Computing (ARC) facility. Available to the Virginia Tech research community, ARC’s mission is to advance computational science, engineering, and technology.

Terascala’s TeraOS HPC storage management software has been a part of the university’s flagship cluster, BlueRidge, since June of this year. BlueRidge is a 408-node Cray CS-300 cluster with 6,528 cores and 27.3 TB of memory systemwide. With a theoretical peak performance of approximately 400 teraflops per second, BlueRidge allows researchers to run massively-parallel simulations and focus on computationally-intensive problems.

“Terascala’s TeraOS provides a truly complete Lustre parallel file system solution,” explained Brian Marshall, computational scientist at Virginia Tech. “The hardware installation and setup were straightforward. The web dashboard provides real-time monitoring and management capabilities, easing the burden of administration.”

According to Marshall, the high performance storage appliance is an integral part of the BlueRidge cluster. “ARC uses Lustre as the parallel scratch filesystem for BlueRidge. Many distributed applications require a large, shared space to write temporary and checkpoint results. Without a high performance, parallel filesystem, file I/O to a single disk would create a crippling bottleneck, resulting in most processors idling. Terascala's Lustre solution removes this bottleneck, allowing applications to take full advantage of the computational power available.”

The Terascala-managed HPC storage appliance works well, allowing Marshall and his team to focus on algorithms and other computational tasks instead of trouble-shooting file systems. “The system is a turnkey appliance that definitely simplifies file system administration,” continued Marshall. “It allows us to spend our time elsewhere–which is extremely valuable.”

In regards to taking advantage of all of the advanced capabilities of their new system, Marshall believes that Virginia Tech is just scratching the surface. “We are still in the process of optimizing our applications for the new file system and are seeing promising, initial results.”

About Terascala

Terascala—the High Performance Computing (HPC) storage management company—has pioneered software that lowers total cost of ownership by managing and optimizing data, performance, and reliability. Savings are quickly realized through minimizing downtime, moving data very fast between scratch and inexpensive backup storage, and optimizing workload throughput. The company’s TeraOS software provides system-level high availability to reduce costly degraded performance and downtime. It integrates workload, network, storage, and file system monitoring, analysis, pre-emptive failure alerts, and fast failover. Terascala’s real-time analysis and phone-home support keeps these highly complex systems up and running at peak performance. Support costs are minimized with no HPC or parallel file system expertise required.

Connect with Terascala

Follow us on Twitter: http://www.twitter.com/terascala_inc
Learn more at http://www.terascala.com

Terascala and the Terascala logo are trademarks or registered trademarks of Terascala, Inc. All other brands, products, or service names may be trademarks or property of their respective holders.

Share article on socal media or email:

View article via:

Pdf Print

Contact Author

Alison Golan
Golan Public Relations
+1 904-230-3369
Email >

Lindsey Angione
Terascala
508-588-1501 218
Email >
Visit website