Data grids, collections, and grid bricks

Abstract

Data grids federate storage resources. They provide a logical name space that can be used to register digital entities, a storage repository abstraction for manipulating data, and a high level abstraction for supporting user-selected interfaces. Data grids can be used to build persistent collections. Data can be stored across multiple types of storage systems with persistent copies kept in archives. Persistent identifiers can be kept in the logical name space. By integrating grid bricks (commodity based disk caches) with archival storage systems, one can assemble a data management environment that supports both interactive access (data picking) and long-term persistent storage. Examples of the creation of interactive data picking environments will be given that integrate grid brick technology with large-scale archives.