DOI QR코드

DOI QR Code

A method for optimizing lifetime prediction of a storage device using the frequency of occurrence of defects in NAND flash memory

낸드 플래시 메모리의 불량 발생빈도를 이용한 저장장치의 수명 예측 최적화 방법

  • Lee, Hyun-Seob (Division of Computer Engineering, Baekseok University)
  • 이현섭 (백석대학교 컴퓨터공학부)
  • Received : 2021.08.26
  • Accepted : 2021.10.16
  • Published : 2021.12.31

Abstract

In computing systems that require high reliability, the method of predicting the lifetime of a storage device is one of the important factors for system management because it can maximize usability as well as data protection. The life of a solid state drive (SSD) that has recently been used as a storage device in several storage systems is linked to the life of the NAND flash memory that constitutes it. Therefore, in a storage system configured using an SSD, a method of accurately and efficiently predicting the lifespan of a NAND flash memory is required. In this paper, a method for optimizing the lifetime prediction of a flash memory-based storage device using the frequency of NAND flash memory failure is proposed. For this, we design a cost matrix to collect the frequency of defects that occur when processing data in units of Drive Writes Per Day (DWPD). In addition, a method of predicting the remaining cost to the slope where the life-long finish occurs using the Gradient Descent method is proposed. Finally, we proved the excellence of the proposed idea when any defect occurs with simulation.

고신뢰성을 요구하는 컴퓨팅 시스템에서 저장장치의 수명예측방법은 데이터 보호뿐만 아니라 활용성을 극대화 할 수 있기 때문에 시스템 관리하기 위한 중요한 요소 중 한 가지이다. 최근 여러 저장시스템에서 저장장치로 사용되고 있는 SSD(Solid State Drive)의 수명은 이를 구성하고 있는 낸드 플래시 메모리의 수명이 실질적인 수명과 연결된다. 따라서 SSD를 이용하여 구성한 저장시스템에서는 낸드 플래시 메모리의 수명을 정확하고 효율적으로 예측하는 방법이 필요하다. 본 논문에서는 낸드 플래시 메모리 불량 발생빈도를 이용하여 플래시 메모리 기반 저장장치의 수명 예측을 최적화할수 있는 방법을 제안한다. 이를 위해 DWPD(Drive Writes Per Day) 단위로 데이터를 처리할 때 발생하는 불량 발생빈도를 수집하기 위한 비용 매트릭스(Cost Metrix)를 설계한다. 그리고 경사하강법(Gradient Descent)을 이용하여 수명의 마감이 발생하는 경사도까지 남은 비용을 예측하는 방법을 제안한다. 마지막으로 시뮬레이션을 통해 임의의 불량이 발생했을 때 제안하는 방법을 통한 수명예측의 우수성을 증명했다.

Keywords

Acknowledgement

이 논문은 2021학년도 백석대학교 학술연구비 지원을 받아 작성되었음

References

  1. S.S.Chae, R.Mativenga, J.Y.Paik, M.Attique, and T.S.Chung, "DSFTL: An efficient FTL for flash memory based storage systems." Electronics Vol.9, No.1, pp.145, 2020. https://doi.org/10.3390/electronics9010145
  2. W.Xie, Y.Chen, and P.C.Poth, "ASA-FTL: An adaptive separation aware flash translation layer for solid state drives," Parallel Computing, Vol.61, pp.3-17, 2017. https://doi.org/10.1016/j.parco.2016.10.006
  3. I.B.Zion, "Key-value FTL over open channel SSD," 12th ACM International Conference on Systems and Storage. pp.192-192, 2020.
  4. S.Kim and Y.Son, "Optimizing Key-Value Stores for Flash-Based SSDs via Key Reshaping," IEEE Access 9, pp.115135-115144, 2021. https://doi.org/10.1109/ACCESS.2021.3105428
  5. J.H.Park, D.J.Park, T.S.Chung, and S.W.Lee, "ARK, Jong-Hyeok, et al. A Crash Recovery Scheme for a Hybrid Mapping FTL in NAND Flash Storage Devices," Electronics, Vol.10, No.3, pp.327, 2021. https://doi.org/10.3390/electronics10030327
  6. Samsung Electronics, Samsung Solid State Drive. http://www.samsung.com/zglobal/business/semicond uctor/support/brochures/support_Flash_SSD.html..
  7. H.S.Lee and D.H.Lee, "An efficient index buffer management scheme for implementing a B-tree on NAND flash memory," Data & Knowledge Engineering. Vol.69, No.9, pp.901-916. 2010. https://doi.org/10.1016/j.datak.2010.03.004
  8. H.S.Lee, H.S.Yun, and D.H.Lee, "HFTL: hybrid flash translation layer based on hot data identification for flash memory." Consumer Electronics, IEEE Transactions. Vol.55, No.4, pp.2005-2011, 2011.
  9. H.S.Lee, S.W.Park and D.H.Lee, "RMSS: an efficient recovery management scheme on NAND flash memory based solid state disk,", IEEE Transactions on Consumer Electronics, Vol.59, No.1, pp.107-112, February 2013. https://doi.org/10.1109/TCE.2013.6490248
  10. Y.Pan, H.Chen, J.Zhao, and Y.Xu, "Lifetime-aware FTL to improve the lifetime and performance of solid-state drives." Future Generation Computer Systems 93, pp.58-67, 2019. https://doi.org/10.1016/j.future.2018.10.011
  11. Y.Lue, Y.Cai, S.Ghose, J.Choi, and O.Mutlu, "WARM: Improving NAND flash memory lifetime with write-hotness aware retention management," Mass Storage Systems and Technologies, pp.1-14, 2015.
  12. J.H.Choi, K.M.Kim, and J.W.Kwak, "WPA: Write Pattern Aware Hybrid Disk Buffer Management for Improving Lifespan of NAND Flash Memory," IEEE Transactions on Consumer Electronics, Vo.66, No.2, pp.193-202., 2020. https://doi.org/10.1109/tce.2020.2981618
  13. B.Kai, X.Hui, X.Qiyou, and Y.Wei, "Schemes for Extending the Lifetime of SSD," TELKOMNIKA Indonesian Journal of Electrical Engineering, Vol.12, No.9, pp6940-6945, 2014.
  14. S.Wang, F.Wu, Z.Lu, Y.Zhau, Q.Xiong, M.Zhang, and C.Xie, "Lifetime adaptive ECC in NAND flash page management. In Design," Automation & Test in Europe Conference & Exhibition (DATE), pp.1253-1556, 2017.
  15. Y.J.Woo, "Diversifying wear index for MLC NAND flash memory to extend the lifetime of SSDs," Embedded Software (EMSOFT), pp1-10., 2013.