Write Amplification Trade-off Analysis in Hybrid Mapping Solid State Drives

Write amplification defines the efficiency of write operation, and it directly affects the Input/Output operations per second (IOPS) and endurance of NAND flash memories. The log-updated write scheme of NAND flash based Solid State Drives will renewal data in different places. Garbage collection will cause extra I/O operations which affects the write amplification factor. Hybrid mapping scheme alleviates the large size of mapping table in page mapping scheme, and the IOPS of storage devices in block mapping scheme, so the write amplification trade-off has great effects on the design of hybrid mapping scheme. In this paper, we propose a model to analysis the write amplification trade-off in hybrid mapping scheme. Moreover, we redesign the greedy garbage collection and measure the performance differences between two algorithms in hybrid mapping scheme. Finally, we evaluate the design of write amplification trade-off based hybrid mapping Solid State Drives, and the algorithms proposed in this paper can resuce the write amplification factor and maintain the endurance at a high level.

[1]  Evangelos Eleftheriou,et al.  Write amplification analysis in flash-based solid state drives , 2009, SYSTOR '09.

[2]  Sang-Won Lee,et al.  A log buffer-based flash translation layer using fully-associative sector translation , 2007, TECS.

[3]  LimSeung-Ho,et al.  An Efficient NAND Flash File System for Flash Memory Storage , 2006 .

[4]  Li Wang,et al.  Research on the Flash Translation Layer Based on Grouping Pages , 2016, 2016 Sixth International Conference on Instrumentation & Measurement, Computer, Communication and Control (IMCCC).

[5]  Yang Yan,et al.  A wear-leveling algorithm based on IO request prediction , 2012, 2012 3rd International Conference on System Science, Engineering Design and Manufacturing Informatization.

[6]  Eun-Sun Cho,et al.  Probability-based static wear-leveling algorithm for block and hybrid-mapping NAND flash memory , 2012, Des. Autom. Embed. Syst..

[7]  Zhou Li,et al.  A Novel NAND Flash FTL for Mass Data Storage Devices Based on Hybrid Address Translation , 2013, 2013 Third International Conference on Intelligent System Design and Engineering Applications.

[8]  KangSooyong,et al.  Performance Trade-Offs in Using NVRAM Write Buffer for Flash Memory-Based Storage Devices , 2009 .

[9]  Kern Koh,et al.  A Demand-Based FTL Scheme Using Dualistic Approach on Data Blocks and Translation Blocks , 2011, 2011 IEEE 17th International Conference on Embedded and Real-Time Computing Systems and Applications.

[10]  Youngjae Kim,et al.  DFTL: a flash translation layer employing demand-based selective caching of page-level address mappings , 2009, ASPLOS.

[11]  Mohammed Azmi Al-Betar,et al.  A survey of techniques for architecting SLC/MLC/TLC hybrid Flash memory–based SSDs , 2018, Concurr. Comput. Pract. Exp..