论文快速发表网

社科类论文科技类论文医学类论文管理类论文教育类论文农林类论文新闻类论文建筑类论文文艺类论文法学类论文

EI Compendex Source List（2022年1月） EI Compendex Source List（2020年1月） EI Compendex Source List（2019年5月） EI Compendex Source List（2018年9月） EI Compendex Source List（2018年5月） EI Compendex Source List（2018年1月）中国科学引文数据库来源期刊列 CSSCI(2017-2018)及扩展期刊目录 2017年4月7日EI检索目录（最新） 2017年3月EI检索目录最新公布北大中文核心期刊目录 SCI期刊（含影响因子）

本站服务项目浅谈青年教师如何开展教科研大学青年教师如何处理教学和科

论文范文

CRIM: Conditional Remapping to Improve the Reliability of Solid-State Drives with Minimizing Lifetime 时间:2018-09-23 23:57   来源:未知   作者:admin   点击: 次 Abstract:Solid-state drive (SSD) becomes popular as the main storage device. However, over time, the reliability of SSD degrades due to bit errors, which poses a serious issue. The periodic remapping (PR) has been suggested to overcome the issue, but it still has a critical weakness as PR increases lifetime loss. Therefore, we propose the conditional remapping invocation method (CRIM) to sustain reliability without lifetime loss. CRIM uses a probability-based threshold to determine the condition of invoking remapping operation. We evaluate the effectiveness of CRIM using the real workload trace data. In our experiments, we show that CRIM can extend a lifetime of SSD more than PR by up to 12.6% to 17.9% of 5-year warranty time. In addition, we show that CRIM can reduce the bit error probability of SSD by up to 73 times in terms of typical bit error rate in comparison with PR.
1. Introduction
  Solid-state drive (SSD) is a high-performance and energy-efficient storage device based on NAND flash memory. SSD has become increasingly popular in recent times, seeing usage from in personal computers to in high-performance servers [1, 2]. In particular, SSD can enhance the performance of cloud storage systems. Cloud storage systems in data centers should consider energy efficiency, performance, and maintenance costs [3, 4]. In large-scale cloud storage system, the replacement costs of storage devices can be concerned. SSDs can be the key solution because their performance is much faster than that of hard disk drives (HDDs), and their cost is much lower than that of static random access memory (SRAM).
However, SSD has reliability problems because NAND flash memory has bit errors. The bit errors degrade the performance and lifetime of SSD [5]. Especially, the degradation of lifetime is a serious problem because it reduces the amount of available blocks which are limited resources. Thus, enhancing lifetime is an important and urgent demand, so many studies have been proposed until recently [6–10].
Many SSDs handle bit errors by applying error correction code (ECC) mechanisms such as BCH [11]. However, ECC does not completely solve the bit error problem because the error correction capability of ECCs is not scalable. For example, BCH-512 can correct 7 bit errors with a 512-length code. BCH-32 k can correct 259 bit errors—the code length is 64 times longer than BCH-512 but the number of correctable errors is only 37 times more. Furthermore, the power consumption and additional space for metadata increase by multiples of 71 and 85, respectively [12].
  After that, more advanced way to handle bit errors has been suggested: applying the periodic remapping (PR) of data blocks [12]. PR periodically replaces error-prone blocks into healthy ones. However, PR has a critical weakness in that PR reduces the overall lifetime of SSDs. Thus, we attempt to find a way to minimize the lifetime loss caused by PR.
  Our key observation is that the cause of lifetime loss is invoking remapping operations too often in PR. As solution, we propose CRIM, the conditional invocation of the remapping operation based on failure likelihood. To reduce lifetime loss, CRIM invokes the remapping operation only when a certain condition is met. The condition is based on the probability of bit error occurrence, and it determines how much a target block is error prone.