SDSC Announces Launch of Academic-based High-performance Cloud Storage System

The San Diego Supercomputer Center (SDSC) at the University of California, San Diego, announced the introduction of an academic-based cloud storage system which has been developed specifically for the academic community. The cloud-based storage system has been set up as a secure and cost-effective storage facility which will enable easy sharing of digital information. 

In a release, Michael Norman, director of SDSC, said that, “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. The SDSC Cloud goes a long way toward meeting federal data sharing requirements, since every data object has a unique URL and could be accessed over the Web.”

The new scalable cloud-based storage system, which was initially planned for UC San Diego's campus Research Cyberinfrastructure (RCI) project, has a disk-based design and incorporates high-speed 10 gigabit Ethernet switching technology to enable super-fast read and write capabilities. Housing 5.5 petabytes capacity (1 petabyte equals 1 quadrillion bytes of storage capacity), SDSC's Cloud system is capable of providing 8 to 10 gigabytes of read rate per second. This initial capacity is however, set to extend and improve as additional nodes and storage are included.

Data which is stored in the new cloud-based system is constantly checked for consistency and is immediately written to multiple independent storage servers. Developed on a performance centric parallel file system architecture, the SDSC Cloud leverages two Arista Networks 7508 switches (configured over multi-chassis link aggregation) which extend 768 total 10 gigabit  Ethernet ports to deliver in excess of 10Tbit/s of IP-based connectivity.

Because the system offers high-bandwidth wide-area network connectivity via multiple 10 GB connections to the Corporation for Education Network Initiatives in California, Energy Sciences Network, and the Extreme Science and Engineering Discovery Environment, massive amounts of data can be easily transferred across the SDSC Cloud.

Users can opt for regular “on-demand” storage for $3.25 a month per 100GB of storage and are not charged for I/O networking or they can opt for a cost-effective “condo” option, allowing them to make long term investment for hardware included in the SDSC Cloud. Furthermore, an option for storing additional copies of their data at UC Berkeley, a partner in the project, will also be offered to users soon. Since each file is given a dedicated URL, the system is primed for sharing of data.

Users can choose between open and private access and because it features the RackSpace Swift (News - Alert) and Amazon S3 APIs, users can access the system via any web browser and mobile devices and have the additional option of writing applications for direct interaction with the SDSC Cloud. SDSC's Cloud provides increased security as users can set access and privacy levels on their own. Furthermore, are aware of the location of their data and the replicated copies at all times.

The system has been designed to collaborate with other storage technologies at the supercomputer center such as the Data Oasis system, used by researchers working on data critical operations over SDSC's Triton, Trestles, and Dash systems.