Supercomputer Center Claims Largest Storage Cloud

By Stuart J. Johnston

What do you call it when an cloud storage system gets so large and fast it would seem to need its own storm warning?

You might want to call it the San Diego Supercomputer Center (SDSC) Cloud at the University of California, San Diego.

The UCSD's SDSC debuted this week what it believes is "the largest academic-based cloud storage system in the U.S., specifically designed for researchers, students, academics, and industry users who require stable, secure, and cost-effective storage and sharing of digital information, including extremely large data sets," according to a university statement.

What's more, it's relatively inexpensive to use -- starting at $3.25 per month for 100 GB of storage.

"We believe that the SDSC Cloud may well revolutionize how data is preserved and shared among researchers, especially massive datasets that are becoming more prevalent in this new era of data-intensive research and computing," Michael Norman, director of the SDSC, said in a statement.

The SDSC Cloud is entirely disk-based and is interconnected by 10 gigabit Ethernet (10GbE) switches. Its current raw storage capacity is 5.5 petabytes but over time, the Cloud is expandable to hundreds of petabytes.

In terms of sustained read rates, at 8 to 10 GB per second, the SDSC Cloud has the capability to read a 250 GB laptop hard drive in under 30 seconds.

"This allows huge amounts of data, such as sky surveys or mapping of the human genome, to be rapidly transported simultaneously to/from the SDSC Cloud," the statement added.

The massive read and write speeds and immediate availability of users' data may well change the way that researchers think about their data.

"We are shifting from the 'write once and read never' model of archival data, to one that says 'if you think your data is important, then it should be readily accessible and shared with the broader community," Richard Moore, SDSC's deputy director, said in a statement.

SDSC Cloud is built on a scalable, open-sourced cloud operating system called OpenStack, which was launched jointly in 2010 by NASA and Rackspace Hosting.

Key to the cloud's architecture, data's durability and security is maintained by the fact that data is instantly written to multiple independent servers, and constantly validated for consistency.

The physical system uses a high-performance parallel file system that runs across two Arista Networks 7508 switches, providing a total of 768 10GbE ports that deliver more than 10 terabits of non-blocking, IP-based connectivity, the SDSC statement said.

Further information regarding SDSC Cloud is available online.

Stuart J. Johnston is a contributing editor at InternetNews.com, the news service of Internet.com, the network for technology professionals. Follow him on Twitter @stuartj1000.

This article was originally published on September 28, 2011