WAPFTL:支持预测机制的负载自适应闪存转换层算法

J4 ›› 2014, Vol. 36 ›› Issue (07): 1238-1243.

WAPFTL:支持预测机制的负载自适应闪存转换层算法

谢徐超，宋振龙，李琼，魏登萍，方健，肖立权

（国防科学技术大学计算机学院，湖南长沙 410073）

收稿日期:2013-08-13 修回日期:2013-11-21 出版日期:2014-07-25 发布日期:2014-07-25
基金资助:
国家863计划资助项目（2012AA01A301）；国家自然科学基金资助项目（61202118）

WAPFTL:A workload adaptive page-level
flash translation layer with prediction

XIE Xuchao,SONG Zhenlong,LI Qiong,WEI Dengping,FANG Jian,XIAO Liquan

(College of Computer,National University of Defense Technology,Changsha 410073,China)

Received:2013-08-13 Revised:2013-11-21 Online:2014-07-25 Published:2014-07-25

摘要/Abstract

摘要：

基于NAND Flash的固态盘凭借其低延迟、低功耗、高可靠性等优点，已经开始应用于企业级服务器和高性能计算领域。针对固态盘相对较差的写性能及使用寿命有限等不足，提出了一种闪存转换层中基于页映射机制的自适应地址映射算法WAPFTL。该算法能够在地址转换过程中预测负载读写特性并自适应地调整地址映射信息缓存的策略。实验结果表明，WAPFTL能够高效协同利用负载的时间局部性和空间局部性，提高地址映射命中率，减少因地址映射而引起的额外写操作次数;同时，有效减少了垃圾回收次数，提高了SSD整体性能。

关键词: NAND Flash, 固态盘, 地址映射, 闪存转换层

Abstract:

NAND Flash based Solid State Drive (SSD) has been utilized in enterprise servers and high performance computing environments due to its various advantages, such as fast access speed, low power consumption, high reliability, etc. Since SSD suffers from poor performance of random writes and limited lifespan, a workload adaptive pagelevel address mapping algorithm, named WAPFTL, is proposed, which can predict the characteristics of incoming requests and dynamically adapt the cache strategy of address mapping information. Experimental results show that WAPFTL can improve the cache hit ratio and reduce the number of expensive extra read/write operations by jointly exploiting the temporal and spatial localities of workloads efficiently. Besides, WAPFTL can also reduce the number of erase operations, thus improving the overall performance of SSD.

Key words: NAND flash;SSD;address mapping;flash translation layer

谢徐超，宋振龙，李琼，魏登萍，方健，肖立权. WAPFTL:支持预测机制的负载自适应闪存转换层算法[J]. J4, 2014, 36(07): 1238-1243.

XIE Xuchao,SONG Zhenlong,LI Qiong,WEI Dengping,FANG Jian,XIAO Liquan. WAPFTL:A workload adaptive page-level
flash translation layer with prediction [J]. J4, 2014, 36(07): 1238-1243.

参考文献

［1］Chung T S, Park D J, Park S, et al. A survey of flash translation layer［J］. Journal of Systems Architecture, 2009, 55(5):332343.
［2］Lim S P, Lee S W, Moon B. Faster FTL for enterpriseclass flash memory SSDs［C］∥Proc of 2010 IEEE International Workshop on Storage Network Architecture and Parallel I/Os (SNAPI), 2010:312.
［3］Wei Q, Gong B, Pathak S, et al. WAFTL:A workload adaptive flash translation layer with data partition［C］∥Proc of 2011 IEEE 27th Symposium on Mass Storage Systems and Technologies (MSST), 2011:112.
［4］Wang C, Wong W F. ADAPT:Efficient workloadsensitive flash management based on adaptation, prediction and aggregation［C］∥Mass Storage Systems and Technologies (MSST),2012 IEEE 28th Symposium on. IEEE,2012:112.
［5］Gupta A, Kim Y, Urgaonkar B. DFTL:A flash translation layer employing demandbased selective caching of pagelevel address mappings［C］∥Proc of the 14th International Conference on Architectural Support for Programming Languages and Operating System,2009:229240.
［6］Karedla R, Love J S, Wherry B G. Caching strategies to improve disk system performance［J］. Computer, 1994, 27(3):3846.
［7］Kim Y, Tauras B, Gupta A, et al. Flashsim:A simulator for NAND flashbased solidstate drives［C］∥Proc of the First IEEE International Conference on Advances in System Simulation, 2009:125131.
［8］Lee S W, Park D J, Chung T S, et al. FAST:A logbuffer based FTL scheme with fully associative sector translation［J］. ACM Transations on Embedded Computing Systems, 2007,6(3):118.
［9］Storage performance council (SPC) storage traces［EB/OL］.［20070601］. http://traces.cs.umass.edu/index.php/Storage/Storage.
［10］Narayanan D, Donnelly A, Rowstron A. Write offloading:Practical power management for enterprise storage［J］. ACM Transactions on Storage (TOS), 2008, 4(3):10.

[1]	龚锐, 石伟, 刘威, 张剑锋, 王蕾. 基于NAND Flash的CPU安全启动设计与实现[J]. 计算机工程与科学, 2022, 44(06): 971-978.
[2]	刘彦，徐骥龙，祝磊. 基于权重堆排序的NAND Flash静态磨损均衡机制[J]. 计算机工程与科学, 2019, 41(02): 207-213.
[3]	罗志勇，王青青，王晴. 无线传感器网络标识解析的系统设计[J]. J4, 2014, 36(01): 51-56.
[4]	张辉1，2，晏益慧1，罗宇1. 大容量NAND Flash文件系统转换层优化设计[J]. J4, 2011, 33(4): 81-85.
[5]	张翔立，翟建东. 碎片影响闪存性能分析与改进[J]. J4, 2011, 33(1): 66-69.
[6]	王顺,戴瑜兴,段小康. 基于FPGA的NAND Flash控制接口电路设计[J]. J4, 2010, 32(7): 80-82.
[7]	吴俊军,刘震宇. 一种基于功能表的高效FTL算法[J]. J4, 2010, 32(11): 128-131.
[8]	王同洋董永帅吴俊军. 基于阈值控制的损耗均衡算法设计[J]. J4, 2008, 30(3): 72-74.
[9]	曹计昌邱鹏张斌. UCard中动态地址映射的实现方法研究[J]. J4, 2006, 28(1): 119-121.