计算机科学 ›› 2017, Vol. 44 ›› Issue (6): 51-56.doi: 10.11896/j.issn.1002-137X.2017.06.008

• 2016 年全国信息存储技术学术年会 • 上一篇    下一篇

一种固态盘的读写性能优化调度方法

朱玥,吴非,熊钦,谢长生   

  1. 武汉光电国家实验室 武汉430074,武汉光电国家实验室 武汉430074;信息存储系统教育部重点实验室 武汉430074,武汉光电国家实验室 武汉430074,武汉光电国家实验室 武汉430074;信息存储系统教育部重点实验室 武汉430074
  • 出版日期:2018-11-13 发布日期:2018-11-13
  • 基金资助:
    本文受国家自然科学基金(61300047),中央高校基本科研业务费(HUST:2016YXMS019)资助

Read-Write Performance Optimization Scheduling Scheme for SSD

ZHU Yue, WU Fei, XIONG Qin and XIE Chang-sheng   

  • Online:2018-11-13 Published:2018-11-13

摘要: 相比于传统机械硬盘,基于NAND Flash的固态盘由于具有非易失性、高性能、低功耗等优点,被广泛应用于数据中心、云计算、在线事务交易等场景。然而,由于NAND Flash中的读操作速度远远快于写操作速度,当读写请求并发执行时,读请求可能被写请求阻塞,从而表现出极大的读延时。在许多以读请求为主的场合,尤其是在线事物交易中(读请求占总请求的比例超过90%),读延时的急剧增加严重影响了系统的整体性能。提出一种读写性能优化调度的策略,通过在闪存转换层之下动态调整读写请求的优先序列,使读性能获得显著的提升。实验中,通过对固态盘仿真器的设计与实现,对读写调度策略的有效性进行了系统的评估。实验结果表明,在该调度策略下,系统中读延时的最大值和平均值均得到了显著的减少,且降幅分别达到了72%和41%。

关键词: 闪存,固态盘,数据调度,队列管理

Abstract: Compared with traditional hard disk drives (HDDs),NAND-Flash-based solid-state drives (SSDs) are non-volatile and can provide better performance as well as lower power consumption.Therefore,they have achieved extensive application in data centers,cloud computing and online transaction trading,etc.However,in NAND Flash memory,the speed of read operation is significantly faster than the write operation.Hence,for a concurrent workload with a mixture of read and write requests,reads may be blocked by writes,which exhibites an enormous read latency.In many read-intensive applications,especially the online transaction trading,in which the proportion of read requests is than 90%,the sharp increase of the read latency influences the overall performance of the system severely.In this paper,we proposed a read-write performance optimization scheduling scheme which achieves remarkable improvement about the read performance by dynamically adjusting the priority sequence of read and write requests beneath the flash translation layer.In the experiment,we designed and built an SSD simulator to evaluate the effectiveness of the scheduling scheme.Experimental results show that by implementing the proposed scheme,the maximum and the average read latency in the system are substantially reduced,with the reduction of 72% and 41%,respectively.

Key words: Flash memory,SSD,Data scheduling,Queue management

[1] QIN Z,WANG Y,LIU D,et al.Real-time flash translation layer for NAND flash memory storage systems [C]∥Proc of IEEE RTAS’12.Piscataway,NJ:IEEE,2012:35-44.
[2] JUNG M,CHOI W,SRIKANTAIAH S,et al.HIOS:A host interface I/O scheduler for solid state disks [J].ACM SIGARCH Computer Architecture News,2014,2(3):289-300.
[3] LEE J,KIM Y,SHIPMAN G M,et al.Preemptible I/O scheduling of garbage collection for solid state drives [J].IEEE Tran-sactions on CADCS,2013,2(2):247-260.
[4] ABDURRAB A R,XIE T,WANG W.DLOOP:A flash translation layer exploiting plane-level parallelism [C]∥Proc of IEEE IPDPS’13.Piscataway,NY:IEEE,2013:908-918.
[5] GAO C,SHI L,ZHAO M,et al.Exploiting parallelism in I/Oscheduling for access conflict minimization in flash-based solid state drives [C]∥ Proc of IEEE MSST’14.Piscataway,NY:IEEE,2014:1-11.
[6] PARK S,SHEN K.FIOS:A fair,efficient flash I/O scheduler [C]∥Proc of USENIX FAST’12.Berkeley,CA:USENIX,2012:1-15.
[7] PARK S,SEO E,SHIN J Y,et al.Exploiting internal paralle-lism of flash-based SSDs [J].Computer Architecture Letters,2010,9(1):9-12.
[8] CHEN F,LEE R,ZHANG X.Essential roles of exploiting internal parallelism of flash memory based solid state drives in high-speed data processing [C]∥Proc of IEEE HPCA’11.Piscata-way,NY:IEEE,2011:266-277.
[9] PARK S,SHEN K.A performance evaluation of scientific I/O workloads on flash-based SSDs[C]∥Proc of IEEE CLUSTER’09.Piscataway,NY:IEEE,2009:1-5.
[10] CAI Y,LUO Y,GHOSE S,et al.Read disturb errors in MLC NAND flash memory:Characterization,mitigation,and recovery [C]∥ Proc of IEEE DSN’15.Piscataway,NY:IEEE,2015:438-449.
[11] YANG M C,CHANG Y H,TSAO C W,et al.Utilization-aware self-tuning design for TLC flash storage devices [J].IEEE Trans.on Very Large Scale Integration Systems,2016,24(10):3132-3144.
[12] BEZ R,CAMERLENGHI E,MODELLI A,et al.Introduction to flash memory [J].Proceedings of the IEEE,2003,1(4):489-502.
[13] JUNG D,CHAE Y H,JO H,et al.A group-based wear-leveling algorithm for large-capacity flash memory storage systems [C]∥Proc of ACM CASES’07.New York,NY:ACM,2007:160-164.
[14] HU Y,JIANG H,FENG D,et al.Exploring and exploiting the multilevel parallelism inside SSDs for improved performance and endurance [J].IEEE Transactions on Computers,2013,2(6):1141-1155.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!