Large-scale storage systems are well served by heterogeneous architectures. Allowing devices of varying characteristics and capabilities to smoothly interact allows a storage system to grow and evolve gracefully. A system comprised of heterogeneous devices also offers unique opportunities for administrators to dictate when and where devices are integrated and utilized based upon their characteristics.

To address these opportunities, we have developed Logan. Logan optimizes the growth of a system by choosing which devices to integrate into the system based on administratively defined policies. Similarly, it maintains and improves system state by allowing administrators to dictate at a high level when and where data should be migrated or rebuilt when a device fails or is decommissioned.

To validate Logan’s decision making abilities, we tested it using a discrete event simulator that tracked the power usage of a Logan-managed storage system guided by an energy-aware policy. Logan was able to improve average power usage of a large-scale system when compared to naıve techniques that expand the system without regard to underlying device characteristics. This improvement is made possible by a high degree of heterogeneity and free space in the system, a common occurrence for large and growing storage systems.