Visit additional Tabor Communication Publications
June 14, 2011
Xyratex is one of oldest and largest storage vendors in the world. The 25-year-old company booked $1.6 billion in revenue last year, yet few people have ever heard of it. That's because it primarily sells its equipment through OEM relationships with companies like IBM, Dell, and NetApp.
Through these partnerships, Xyratex has been able to penetrate deeply into the enterprise storage space. In 2010, they shipped over 1.5 exabytes of external storage and claimed 19 percent of enterprise storage capacity. According to IDC, the company is the largest OEM disk storage system provider in the world.
And thanks to the Xyratex's broad market reach, a significant chunk of this storage ends up in high performance computing environments -- an estimated 375 petabytes in 2010. But in November of last year, Xyratex made a more strategic HPC play by acquiring ClusterStor, a startup led by Lustre inventor Peter Braam.
The acquisition brought Braam, as well as Lustre architect Peter Bojanic, to Xyratex to bootstrap the company's Lustre development team. Prior to joining Xyratex, Bojanic had led the Lustre group at Oracle, the company that inherited the file system through its 2009 acquisition of Sun Microsystems. Sun itself had obtained the Lustre technology through its 2007 acquisition of Cluster File Systems, Braam's and Bojanic's original Lustre venture.
The reunion of Lustre expertise and technology at Xyratex bore its first commercial fruit this week. On Tuesday, the company announced ClusterStor 3000, a rack-scale Lustre storage solution purpose-built for high performance computing. Like Xyratex's other storage offerings, ClusterStor 3000 will be available through OEMs -- as yet unnamed.
According to Xyratex, single rack node versions of ClusterStor 3000 are currently under trial at the University of Florida and University of Cambridge, with additional beta sites pending in oil & gas, research, life sciences, and government. First commercial shipments are expected in the fourth quarter of the year.
HPCwire asked Xyratex's director of Strategic Business Development Ken Claffey to describe the new offering, its value proposition, and how it's positioning the product in the HPC storage landscape.
HPCwire: How did Xyratex come to the HPC business?
Ken Claffey: Xyratex is the largest OEM supplier of storage systems in the world, and while many may not recognize the name, most HPC data centers will have storage system based on Xyratex technology operating in their datacenters today. We pioneered the high density platforms that have become the cornerstone of HPC storage when we delivered our first high density storage platform some seven years ago.
In addition we acquired ClusterStor last year and have expanded the development team to nearly 150 developers since then. ClusterStor was focused on clustered file systems and their application in HPC.
HPCwire: There are other storage clusters on the market that incorporate Lustre. Why are you calling the ClusterStor 3000 the "first true Lustre storage appliance?"
Claffey: We have changed that positioning slightly upon getting feedback from the analyst community. However, the ClusterStor solution is different from other so called appliances through the level of integration it incorporates both at a hardware and software level, from the embedded firmware and diagnostic software within the platform all the way up to managing the entire storage cluster through a single management interface. All key technologies are integrated into a single ClusterStor system, providing the best in scale-out performance, reliability and ease of management.
HPCwire: What kind of storage density and the performance-per-watt does it deliver?
Claffey: The ClusterStor 3000 can support up to 672 3.5-inch drives in a single rack, delivering over 2 petabytes raw and 1.5 petabytes useable capacity per rack -- the highest density in the industry. It provides over 20 gigabytes per second of file system performance per rack, with linear performance scaling as new racks and systems are added.
This is delivered through the ClusterStor3000 unique Scale Out Storage Architecture, that integrates object storage server, RAID and storage capacity within in a singular storage cluster building block called a scalable storage unit (SSU). Therefore, the ClusterStor 3000 can provide 1.5 petabytes of useable RAID6 protected capacity with more than 20 gigabytes per second performance all within a 20KW power envelope. The ClusterStor 3000 delivers the best capacity and performance per watt available.
HPCwire: What else is unique about the technology?
Claffey: The system has been designed from the ground up to be optimized for performance and density. But what it also does through the ClusterStor Manager is alleviate the pain points so often associated with HPC storage and Lustre-based HPC storage in particular; the challenges of installation, of tuning performance and stability; and taming the complexity of managing petabytes of data. It is in these areas where the ClusterStor 3000 really is groundbreaking.
HPCwire: What level of the HPC storage market are you targeting -- large multi-petabyte installations for top-end supercomputers, large and mid-sized cluster installations, or both?
Claffey: The ClusterStor 3000 is targeted at petabyte and above installations, it can however scale from hundreds of terabytes to 30 petabytes and from 2.5 gigabytes per second up to a terabyte per second.
HPCwire: Who do you see as your competition for this product?
Claffey: Our competition will be other OEM suppliers such as DataDirect Networks.
HPCwire: What can you say about your OEM partnerships at this point?
Claffey: All we can say is that we are engaged in qualification with multiple OEMs and that we will be announcing partnerships later this year.
HPCwire: What's on the drawing board for the ClusterStor product line?
Claffey: We have a robust roadmap planned for both the ClusterStor 3000 product line and ongoing enhancements to the Lustre roadmap working in conjunction with the broader community. We will announce these new features as they become available.
May 22, 2013 |
At some point in the not-too-distant future, building powerful, miniature computing systems will be considered a hobby for high schoolers, just as robotics or even Lego-building are today. That could be made possible through recent advancements made with the Raspberry Pi computers.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 15, 2013 |
Supercomputers at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) have worked on important computational problems such as collapse of the atomic state, the optimization of chemical catalysts, and now modeling popping bubbles.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 09, 2013 |
The Japanese government has revealed its plans to best its previous K Computer efforts with what they hope will be the first exascale system...
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.