相較於單層單元（Single-Level Cell，SLC）快閃記憶體，由於多層單元（Multi-Level Cell，MLC）快閃記憶體具有較低的成本優勢及較高的儲存密度，因此市面上大部分的固態硬碟（Solid-State Drives，SSD）皆採用多層單元快閃記憶體作為主要的儲存媒介。不幸的是，多層單元快閃記憶體在可靠性及存取速度上都不及單層單元快閃記憶。其中一種可行的解決方式是引進冗餘磁碟陣列（Redundant Array of Independent Disks，RAID）的概念，來建構一個具高可靠性及高效能的大型快閃記憶體儲存系統。然而，為了實現冗餘磁碟陣列所需的奇偶校驗數據（parity data）將產生大量額外的寫入與更新動作，這使得儲存系統的效能及空間利用率更加惡化。為減少因資料更新而引發奇偶校驗數據的反覆更新，並使資料防護的機制，更靈活、更節省空間，我們針對冗餘固態硬碟陣列提出一個可調式的可靠性架構，在這個架構下，使用者可以根據資料的重要性來設定不同的可靠層級，不同的可靠層級將提供不同層度的資料回復能力。使用者也可以根據需要，動態地改變資料的可靠層級。相對於之前的研究，我們的機制不僅能夠減少大量奇偶校驗數據產生所帶來的存取成本，也兼顧了資料的可靠性，因此更適用於以快閃記憶體為媒介的儲存系統。

Since MLC flash memory is superior to SLC flash memory in terms of cost and density, most solid-state drives (SSDs) use MLC flash memory as major storage medium. Unfortunately, MLC flash memory has relatively poor reliability and slow access. One feasible solution is to adopt the notion of RAID to construct a large-scale flash memory storage system with high reliability and high performance. However, parity data introduced by RAID structure incur extra writes and updates that could further deteriorate performance and space utilization. To reduce repeated updates of parity data incurred by data updates as well as to make data recovery scheme flexible and space efficient, we propose a configurable reliability framework for SSD-based RAID. Under our framework, users can configure different reliability level for their data based on data importance. With different reliability levels, capability of recovery would differ accordingly. Users could also dynamically change the reliability level of their data whenever needed. Different from prior researches, our framework could not only greatly reduce the overhead incurred by parity data but also guarantee the data reliability. Thus, the proposed framework could be applied in flash memory based storage systems.

[1] Soraya Zertal, “A Reliability Enhancing Mechanism for a Large Flash Embedded Satellite Storage System,” ICONS 08 Third International Conference on Systems, 2008.
[2] K. Greenan, D.D.E. Long, E.L. Miller, T. Schwarz, and A. Wildani, “Building Flexible, Fault-Tolerant Flash-Based Storage Systems,” Proc. Fifth Workshop Hot Topics in System Dependability (HotDep ’09), 2009.
[3] Y. Lee, S. Jung, and Y.H. Song, “FRA: A Flash-Aware Redundancy Array of Flash Storage Devices,” IEEE/ACM Int’l Conf. Hardware/Software Codesign and System Synthesis (CODES+ISSS ’09), pp. 163-172, 2009.
[4] Soojun Im, and Dongkun Shin, “Flash-Aware RAID Techniques for Dependable and High-Performance Flash Memory SSD” IEEE Transactions on Computers, Vol. 60, No. 1, Jan. 2011.
[5] Bo Mao, Hong Jiang, Dan Feng, Suzhen Wu, Jianxi Chen, Lingfang Zeng and Lei Tian, “HPDA: A Hybrid Parity-based Disk Array for Enhanced Performance and Reliability” Parallel & Distributed Processing (IPDPS), IEEE International Symposium on 19-23 April 2010.
[6] J.-U. Kang, C. P. J.-S. Kim, H. Park, and J. Lee. “A multichannel architecture for high-performance NAND Flash-based storage system” In Journal of Systems Architecture, pp.644-658, 2007.
[7] Ji-Yong Shin, Zenglin Xia, Ning-Yi Xu, Rui Gao, Xiongfei Cai, Seungryoul Maeng, Feng-Hsiung Hsu “ FTL design exploration in reconfigurable high-performance SSD for server applications” ICS 2009, pp.338-349
[8] Understanding the Flash Translation Layer (FTL) Specification. Technical report, Intel Corporation, Dec 1998.
[9] Flash-memory Translation Layer for NAND Flash (NFTL). Technical report, M-Systems, 1998.
[10] Samsung K9LBG08U0M (v1.0) - 32 GB DDP MLC Data Sheet. Technical report, S. Electronics.
[11] J. W. Hsieh, Y. L. Tsai, T. W. Kuo, and T. L. Lee. Configurable Flash-Memory Management: Performance Versus Overheads. IEEE Transactions on Computers, 57(11):1571-1583, November 2008.

[12] S. W. Lee, D. J. Park, T. S. Chung, D. H. Lee, S. Park, and H. J. Song. A Log Buffer-Based Flash Translation Layer Using Fully-Associative Sector Translation. In ACM Transactions on Embedded Computing Systems, Vol. 6, No. 3, Article 18, July 2007.
[13] J. Kim, J. M. Kim, S. H. Noh, S. L. Min, and Y. Cho. A Space-Efficient Flash Translation Layer for Compact Flash Systems. In IEEE Transactions on Consumer Electronics, Vol. 48, No. 2, pp.366-375, May 2002.
[14] P. M. Chen, E. K. Lee, G. A. Gibson, R. H. Katz, and D. A. Patterson. RAID: high-performance, reliable secondary storage. In ACM Computing Surveys, pp.145-185, 1994.
[15] D. Stodolsky, G. Gibson, and M. Holland. Parity logging overcoming the small write problem in redundant disk arrays. In In ISCA-09: Proc. of the International Symposium on Computer Architecture, pp.64-75, 1993.
[16] J. Menon, J. Roche, and J. Kasson. Floating parity and data disk arrays. Journal of Parallel and Distributed Computing. In ISCA-09: Proc. of the International Symposium on Computer Architecture, pp.129-139, 1993.
[17] OLTP Application I/O. UMass Trace Repository.
Http: //traces.cs.umass.edu/index.php/Storage/Storage.
[18] Samsung MMCRE64G5MXP-0VB (v1.3) -64GB DDP MLC Data Sheet. Technical report, S. Electronics.