ACAR: an adaptive cost aware cache replacement approach for flash memory

  • Authors:
  • Yanfei Lv;Xuexuan Chen;Bin Cui

  • Affiliations:
  • School of Electronics Engineering and Computer Science, Peking University, Key Lab of High Confidence Software Technologies, Ministry of Education, Peking University;School of Electronics Engineering and Computer Science, Peking University, Key Lab of High Confidence Software Technologies, Ministry of Education, Peking University;School of Electronics Engineering and Computer Science, Key Lab of High Confidence Software Technologies, Ministry of Education, Peking University

  • Venue:
  • WAIM'10 Proceedings of the 11th international conference on Web-age information management
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Flash memory has been gaining more popularity as a substitution for magnetic disk. However, due to asymmetric IO latency, cache management policy needs to be reconsidered in systems equipped with flash. A novel buffer replacement approach named ACAR, which stands for Adaptive Cost Aware cache Replacement, is proposed in this paper to address this problem. Taking operation cost into consideration, ACAR allocates two pools for clean and dirty pages separately. In addition, dynamical pool size tuning is also performed according to IO pattern evolvement. Furthermore, hot data recognition capacity is realized in an enhanced version of ACAR. Experiments with artificial and real IO traces demonstrate ACAR outperforms the state-of-the-art cache replacement strategies.