NASA Ames Research Center Relies on Active Archive

The Active Archive Alliance announced that the NASA Ames Research Center is utilizing an active archive to manage and store its high volume data.

Ames is a high-performance computing site for data and provides NASA with advancements in entry, descent and landing technologies; IT; next-generation aviation improvements; astrobiology; airborne sciences; and small satellite programs. It generates approximately 2PB a month of data related to its research and simulation programs.

Ames was an early adopter of active archiving, and has continued to deploy the latest active archive solutions to preserve the massive data volumes it generates and to ensure that its stored data remains secure, reliable and healthy. Its current active archive is architected using Spectra Logic’s enterprise tape libraries integrated with the SGI DMF tiered storage virtualization system. With DMF, all storage, whether disk or tape, is perceived by the users and applications as one very large storage pool. In addition, all data whether on tape or disk, is always online and easy to access.

“The active archive solution allows us to reduce cost through the use of cost-effective tape media for long term data retention in place of disk drives, while maintaining reliability plus the ability to easily retrieve data. Tape is energy efficient resulting in overall energy savings,” said Davin Chan, HPC technical director for CSC supporting NASA Ames Research Center.

Ames has an average of 1,000 users saving and accessing data on a regular basis. Due to the critical nature of its research, all data are stored indefinitely until the users decide to delete it. Approximately 1PB of Ames data is cached on disk at any given time. However, this data is only temporarily stored on disk arrays. When the system reaches 80% capacity, data that has not been recently accessed is automatically migrated to tape.

With the active archive, this migration requires no user interaction. All files still appear to users and applications exactly where they were, no matter where the system may place the files according to data management policies.

“The active archive provides our researchers with fast, online access to their data and provides an effective method to easily expand the storage capacity of the system as storage needs grow,” added Chan.

Case study