面向可靠性的NAND闪存存储系统元数据映射策略

107 浏览量更新于2024-08-27 收藏 931KB PDF 举报

本文档探讨了"面向可靠性的NAND闪存存储系统地址映射策略"（A Reliability-Aware Address Mapping Strategy for NAND Flash Memory Storage Systems）。随着NAND闪存密度的不断提高，其比特错误率也随之上升，这使得传统的错误纠正代码（ECC）在处理多比特错误时的能力大大削弱。由于闪存通常用于存储文件系统元数据和页面映射信息，单个物理页的故障可能会意外且严重地影响整个系统的功能。研究者们提出了Meta-Cure，这是一个集成了硬件和文件系统接口的创新解决方案。Meta-Cure的目标是透明地保护在多比特故障下的元数据，确保存储的稳定性。该策略利用内置的ECC和数据复制技术来保护包含关键数据，如文件系统元数据的页面。通过在运行时形成冗余对，并将它们分布在不同的物理页面上，Meta-Cure旨在提供额外的保护层，抵御潜在的硬件故障。 Meta-Cure的设计重点在于减少对现有文件系统的改动，它能够在不影响正常操作的同时增强系统的可靠性。它的实现方法包括实时检测和管理冗余页面，以及在出现故障时自动切换到备用副本，从而最小化数据丢失和系统中断的风险。此外，论文还可能讨论了Meta-Cure在系统性能、能耗和存储效率方面的权衡，以及与传统方法相比的优势和挑战。这篇研究论文对于面临闪存可靠性问题的现代计算机系统设计者和工程师具有重要意义，提供了针对NAND闪存存储系统的一种有效且透明的故障防护策略，有助于提升存储系统的整体可靠性，尤其是在数据密集型应用环境中。

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, VOL. 33, NO. 11, NOVEMBER 2014 1623

A Reliability-Aware Address Mapping Strategy for

NAND Flash Memory Storage Systems

Yi Wang, Min Huang, Zili Shao, Henry C. B. Chan, Luis Angel D. Bathen,

and Nikil D. Dutt, Fellow, IEEE

Abstract—The increasing density of NAND ﬂash memory leads

to a dramatic increase in the bit error rate of ﬂash, which greatly

reduces the ability of error correcting codes (ECC) to handle

multibit errors.

NAND ﬂash memory is normally used to store

the ﬁle system metadata and page mapping information. Thus,

a broken physical page containing metadata may cause an unin-

tended and severe change in functionality of the entire ﬂash.

This paper presents Meta-Cure, a novel hardware and ﬁle system

interface that transparently protects metadata in the presence of

multibit faults. Meta-Cure exploits built-in ECC and replication

in order to protect pages containing critical data, such as ﬁle

system metadata. Redundant pairs are formed at run time and

distributed to different physical pages to protect against failures.

Meta-Cure requires no changes to the ﬁle system, on-chip hier-

archy, or hardware implementation of ﬂash memory chip. We

evaluate Meta-Cure under a real-embedded platform using a

variety of I/O traces. The evaluation platform adopts dual ARM

Cortex A9 processor cores with 64 Gb

NAND ﬂash memory. We

have evaluated the effectiveness of Meta-Cure on the new tech-

nology ﬁle system ﬁle system. Experimental results show that

the proposed technique can reduce uncorrectable page errors by

70.38% with less than 7.86% time overhead in comparison with

conventional error correction techniques.

Manuscript received November 27, 2013; revised January 30, 2014

and April 7, 2014; accepted July 28, 2014. Date of current version

October 16, 2014. This work was supported in part by the grants from NSF

award CCF-1029783 (Variability Expedition), in part by the Research Grants

Council of the Hong Kong Special Administrative Region, China, under Grant

GRF 15213814, in part by the Germany/Hong Kong Joint Research Scheme

sponsored by the Research Grants Council of Hong Kong and the Germany

Academic Exchange Service of Germany (Reference G_HK021/12), in part by

the National Natural Science Foundation of China under Project 61272103 and

Project 61373049, in part by National 863 Program 2013AA013202, in part by

the Hong Kong Polytechnic University under Grant 4-ZZD7, Grant G-YK24,

Grant G-YM10, and Grant G-YN36, and in part by Academician Workstation

Construction Projects in Guangdong Province (2012B090500020). This ver-

sion is a revised version. A preliminary version of this paper appears in

Proceedings of the 49th IEEE/ACM Annual Design Automation Conference

(DAC’12) [41]. This paper was recommended by Associate Editor J. Henkel.

Y. Wang is with the Guangdong Province Key Laboratory of Popular High

Performance Computers, Shenzhen University, Shenzhen 518060, China, and

also with the Department of Computing, Hong Kong Polytechnic University,

Hong Kong (e-mail: csywang@comp.polyu.edu.hk).

M. Huang is with Automatic Test and Control Institute, Harbin Institute of

Technology, Harbin 150001, China (e-mail: jimmyhuanghit@gmail.com).

Z. Shao is with Embedded Systems and CPS Laboratory, Department

of Computing, Hong Kong Polytechnic University, Hong Kong (e-mail:

cszlshao@comp.polyu.edu.hk).

H. C. B. Chan is with the Department of Computing, Hong Kong

Polytechnic University, Hong Kong (e-mail: cshchan@comp.polyu.edu.hk).

L. A. D. Bathen is with the Programming Systems Laboratory,

Intel Corporation, Santa Clara, CA 95052-8119 USA (e-mail:

danny.bathen@intel.com).

N. D. Dutt is with the Center for Embedded Computer Systems, University

of California, Irvine, Irvine, CA 92697 USA (e-mail: dutt@uci.edu).

Digital Object Identiﬁer 10.1109/TCAD.2014.2347929

Index Terms—Error correcting codes (ECC), memory man-

agement, metadata,

NAND ﬂash memory, redundancy, reliability.

I. INTRODUCTION

URING past decades, the capacity of NAND ﬂash mem-

ory has been increasing dramatically. The dramatic

improvements in capacity and price of

NAND ﬂash mem-

ory have been driven by aggressive shrinking of process

geometry and the increase of the number of bits stored in

each memory cell. However, the increasing density of ﬂash

causes severe reliability issues. Previous studies have shown

that most multilevel cell (MLC) ﬂash chips experience a sharp

increase in bit error rate after a number of reprograms [1]. With

the trend of increasing capacity of ﬂash memory, the bit error

rate will become even worse because of closer voltage levels

assigned to consecutive logic states.

Error correcting codes (ECC) are widely used in

NAND

ﬂash memory to ensure data integrity. ECC can effectively

protect data when the bit error rate is relatively low (e.g.,

SECDED ECC can provide single-error correction and double-

error detection). Since the strength of ECC is predeﬁned and

ﬁxed by chip manufacturer, with the wear of ﬂash, ECC’s

ability to handle multibit errors beyond what it was originally

designed for greatly diminishes over the lifetime of the ﬂash.

This poses a threat to the integrity of metadata (e.g., ﬁle system

metadata, page mappings) stored in ﬂash [2]. If a ﬂash mem-

ory page contains metadata, the data corruption of the page is

very serious, as it may cause an unintended change in function-

ality of the entire ﬂash. While ECC-based approaches improve

the reliability, this paper presents Meta-Cure, a replication-

based approach that signiﬁcantly improves ﬂash reliability and

thus extends the lifetime of ﬂash. Meta-Cure aims to design

a strategy to enhance the built-in ECC and to provide a more

reliable

NAND ﬂash memory storage system.

In order to provide a reliable ﬂash memory storage system,

the primary objective is to ensure the integrity of metadata

stored in ﬂash while considering the distinct characteristics

of ﬂash memory (i.e., limited erase count of ﬂash memory

cells and “out-of-place update”). Most existing approaches

aim to minimize either the amount of critical data or/and

the number of block erase counts [3], [4]. Reliability is

enhanced through the modiﬁcation of different components

NAND ﬂash architecture, such as ﬁle system [5], [6],

parallelism of solid-state drive (SSD) [7], [8], ﬂash trans-

lation layer (FTL) [9], [10], or the optimization of garbage

0278-0070

 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.

See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

下载后可阅读完整内容，剩余8页未读，立即下载

weixin_38499349

粉丝: 2
资源: 961

面向可靠性的NAND闪存存储系统元数据映射策略

ECC Options for Improving NAND Flash Memory Reliability

A Reliability-Augmented Particle Filter for Magnetic Fingerprinting Based Indoor Localization on Smartphone

nd063-c2-design-for-availability-resilience-reliability-replacement-project-starter-template

Energy-Aware Fault-Tolerant Scheduling Under Reliability and Time Constraints in Heterogeneous Systems

Towards-enhancing-grid-reliability-A-framework-to-forecast-electricity-demand-growth

Blockchain Architecture Reliability-Based Measurement for Circuit Unit Importance

ASTM E3159 - 21 Standard Guide for General Reliability - 完整英文版（1

site-reliability-engineer-handbook

Software-Reliability-Prediction-Using-Ensemble-Learning

reliability-paper

最新资源