NCAR to collaborate with San Diego Supercomputer Center on replication of critical data

By Staff
07/21/2006 - 12:00am


   Data cartridges in NCAR's Mass Storage System     Three data silos in NCAR's Mass Storage System
  The Mass Storage System (MSS) at NCAR is a storehouse of irreplaceable research data, digitally archived on 45,000 tape cartridges. In June 2006, NCAR signed a Memorandum of Understanding with the San Diego Supercomputer Center that provides a mechanism by which the two sites can back up each other’s critical data.   NCAR’s MSS contains five enormous data “silos” that house data used by geoscientists around the world for long-range and long-term research. This year, 100 terabytes of these data will be replicated in a similar storage system at San Diego Supercomputer Center. In turn, 100 terabytes of storage space in the NCAR MSS will be devoted to replication of SDSC data.

Last month, the National Center for Atmospheric Research (NCAR) and the San Diego Supercomputer Center (SDSC) signed an important agreement to provide replication and storage for each other’s critical datasets. The agreement formalizes a partnership that will ensure the reliable, long-term preservation of scientific data vital to the missions of both institutions.

In FY2006, NCAR will make 100 terabytes of archival storage space available for replication of SDSC data, while SDSC will do the same for NCAR. The amount of data storage available at each site under the agreement will increase annually by 50 terabytes, reaching 300 terabytes by 2010. The amount can increase by mutual agreement of both sites.

“The direct benefit to NCAR is that we’ll be able to store crucial scientific datasets offsite for business continuity purposes—something we’ve been planning for several years,” says Tom Bettge, deputy director of Operations and Services for NCAR’s Scientific Computing Division. “In the event of an unexpected disaster, critical data on NCAR’s Mass Storage System would be preserved.”

“San Diego is in the same position as NCAR,” he adds. “The SDSC storage silos are at one site, and a disaster could cause the loss of a significant amount of data. This agreement provides a near no-cost, temporary solution for what’s called geographical data replication—the duplication of data at different sites.”

Some of the first NCAR data to be stored at San Diego will be portions of NCAR’s Research Data Archive, which is managed and curated by the Scientific Computing Division. The Research Data Archive contains precious historic records and data from satellites and field experiments, as well as output from global climate-simulation models, mesoscale weather models, and other Earth science models.

“This collaboration will be one of NCAR’s first tangible uses of the TeraGrid,” Bettge says. “Since NCAR and SDSC are now TeraGrid partners, we’ll use the TeraGrid servers at both locations and the 10-gigabit network to transfer data back and forth.” NCAR joined the TeraGrid, the nation’s most comprehensive and advanced infrastructure for open scientific research, in June. News release

“We look forward to developing future similar programs with our other TeraGrid partners,” he notes.

NCAR's Tom Bettge   
NCAR's Tom Bettge is working with colleagues from the San Diego Supercomputer Center on a partnership for data sharing and preservation.  

As set forth in the NCAR/SDSC Memorandum of Understanding, the two institutions will work together to provide access to storage, implement networking procedures, share software tools, create documentation, and conduct random tests of data retrieval from both sites, as well as develop mechanisms to track and catalog data. They will also jointly sponsor and participate in a workshop on data integrity and security.

The data replication effort is a pilot project of the Chronopolis Consortium, a partnership that includes NCAR, SDSC, the University of Maryland, and the University of California Library System. Chronopolis aims to organize, preserve, and make accessible the increasing number of digital holdings that represent vital intellectual assets—many of which, like NCAR’s Research Data Archive, are irreplaceable.

Bettge points out that the SDSC data replication of NCAR data will be available only by invitation and not to all NCAR users. He also emphasizes that the data storage will be in “dark” archives. (Dark archives are inaccessible to the public, preserve data that are available elsewhere, and entail minimal transactions to access the data.)—Lynda Lester

Photos: Lynda Lester, NCAR/CISL