The system will provide the US National Oceanographic and Atmospheric Administration with improved access to large archives of active data at the Geophysical Fluid Dynamics Laboratory
SGI - the technical computing company, announced that the US National Oceanic and Atmospheric Administration (NOAA) has chosen the SGI DMF storage virtualization solution to support NOAA's data archiving requirements at the Geophysical Fluid Dynamics Laboratory (GFDL).
NOAA is a federal agency focused on examining the state of the oceans and atmosphere, while GFDL is the arm that develops and uses mathematical models and computer simulations to improve understanding and prediction of behaviors within the atmosphere, oceans, and weather.
Both entities work under the US Department of Commerce (DOC) to promote expert assessments of local and global climate change through research and computer mathematical models.
Israel Weinman, director of SGI operations at TNN Communications, SGI's representatives in Israel, says that SGI DMF will provide a hierarchical storage management system required to handle the organization's large and active archives, and will allow end users transparent access for critical climate research.
GFDL transfers up to 100 terabytes (100,000 gigabytes) of data per day to and from the archive managed by DMF, making it one of the most active data repositories in the world. With 30 petabytes (30 million billion bytes), it is also the largest database in the world managed by SGI DMF.
GFDL anticipates that its data warehouse will grow to 90 petabytes in the next three years, and Israel Weinman says that the DMF will also address these projected growth rates.
"We selected SGI for this significant deployment of a storage solution, based on the company's proven history of high performance, and the transparent access that the parallel DMF will provide to GFDL researchers," said Steve Baxter, program manager at NOAA.
Israel Vinman points out that SGI DMF automatically creates and manages a tiered virtual storage environment to reduce equipment and operating costs, improve service level and reduce risks.
Parallel DMF provides the security, cost and transparent access benefits of storage systems, with unprecedented scalability.
DMF regularly monitors based on user criteria, and automatically transfers data between storage resources with different cost and performance characteristics. Only the most critical or urgent data resides on high-performance storage media, while less critical data is automatically moved to secondary storage media. The data is always online and available to applications and end users, regardless of the actual location of the storage.
"The parallel DMF is able to support the largest and most active archives in the world, such as those of GFDL," says Weinman. "The new capabilities make it suitable storage to meet the demanding research, storage and support requirements of GFDL".
