Dynamic Data Reallocation in Hybrid Disk Arrays

  • Authors:
  • Tao Xie;Yao Sun

  • Affiliations:
  • University of Alabama, Tuscaloosa;Teradata Corporation, San Diego

  • Venue:
  • IEEE Transactions on Parallel and Distributed Systems
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Current disk arrays consist purely of hard disk drives, which normally provide huge storage capacities with low cost and high throughput for data-intensive applications. Nevertheless, they have some inherent disadvantages such as long access latencies and energy inefficiency due to their build-in mechanical mechanisms. Flash-memory-based solid state disks, on the other hand, although currently more expensive and inadequate in write cycles, offer much faster random read accesses and are much more robust and energy efficient. To combine the complementary merits of hard disks and flash disks, in this paper, we propose a hybrid disk array architecture named hybrid disk storage (HIT) for data-intensive applications. Next, a dynamic data redistribution strategy called performance, energy, and reliability balanced (PEARL), which can periodically redistribute data between flash disks and hard disks to adapt to the changing data access patterns, is developed on top of the HIT architecture. Comprehensive simulations using real-life traces demonstrate that compared with existing data placement techniques, PEARL exhibits its strength in both performance and energy consumption without impairing flash disk reliability.