SFP: Smart File-Aware Prefetching for Flash based Storage Systems

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

Currently, most of the Flash-based storage systems reduce the performance gap between the main memory and storage by data prefetching. However, conventional prefetching techniques perform well on hard disk drives but have limited effectiveness and efficiency on Flash. It is because the complicate data access patterns in modern systems have not been well considered. In this paper, we propose SFP, a smart file-aware prefetching scheme for Flash-based storage systems. SFP demonstrates that prefetching accuracy and efficiency can be improved comprehensively in a file-aware approach. Furthermore, three schemes are proposed: file access pattern learning, dynamic window-based file prefetching, and learning model size optimization. Experiments on the real server show that SFP reduces the access latency by up to 40% compared with the state-of-the-art with low memory and computation cost.

Original languageEnglish
Title of host publicationGLSVLSI 2021 - Proceedings of the 2021 Great Lakes Symposium on VLSI
PublisherAssociation for Computing Machinery
Pages45-50
Number of pages6
ISBN (Electronic)9781450383936
DOIs
StatePublished - 22 Jun 2021
Event31st Great Lakes Symposium on VLSI, GLSVLSI 2021 - Virtual, Online, United States
Duration: 22 Jun 202125 Jun 2021

Publication series

NameProceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

Conference

Conference31st Great Lakes Symposium on VLSI, GLSVLSI 2021
Country/TerritoryUnited States
CityVirtual, Online
Period22/06/2125/06/21

Keywords

  • file prefetching
  • flash
  • storage system

Fingerprint

Dive into the research topics of 'SFP: Smart File-Aware Prefetching for Flash based Storage Systems'. Together they form a unique fingerprint.

Cite this