ABSTRACT: As the capacities of hard drives grow, traditional RAID systems spend an ever increasing amount of time recovering from failed drives. When systems consist of 100’s of drives, reconstruction time becomes a major concern. These large systems also require a substantial amount of knowledge to configure and maintain, which adds to the burden of administrators. Is it possible to rebuild a failed 2TB drive in two hours instead of two weeks? Is it possible to relieve the administrator from having to deal with RAID levels, RAID groups, and Hot Spares? By uniformly distributing RAID data across a large set of drives, the answer to both of these questions is “Yes”. When a single drive failure can significantly impact performance for weeks, it’s time to consider the next generation of data protection.