Device-Aware Cache Management based on Adaptive Replacement

Heterogeneous devices have been adopted widely in mobile storage systems because a combination of such devices can supply a synergistically useful storage solution by taking advantage of each device. In heterogeneous storage systems there have been several researches for enhancing I/O performance by devising proper buffer cache management algorithms. This paper presents a novel device-aware buffer cache management algorithm by employing adaptive cache replacement called ARC, considering both I/O cost per device and workload patterns in mobile computing systems with a heterogeneous storage pair of a hard disk and a NAND flash memory. The proposed algorithm seeks to enhance an existing device-aware cache partitioning technique. Extensive simulations show that the proposed algorithm reduces the total I/O cost over the existing buffer cache algorithms on typical mobile traces. Key-Words: heterogeneous mobile storage, performance optimization, device-aware cache management, dynamic cache partitioning, workload-aware management, adaptive cache replacement