Mitigating Negative Impacts of Read Disturb in SSDs

Mitigating Negative Impacts of Read Disturb in SSDs
复制标题

减轻 SSD 中读取干扰的负面影响

DOI:
10.1145/3410332
复制
发表时间:
2020-09
影响因子:
1.4
通讯作者:
Yutaka Ishikawa
Yutaka Ishikawa
中科院分区:
计算机科学4区
文献类型:
--
作者:
Jun Li;Bowen Huang;Zhibing Sha;Zhigang Cai;Jianwei Liao;Balazs Gerofi;Yutaka Ishikawa

文献摘要

参考文献

被引文献

相似文献

读干扰是固态驱动器(SSD)中的电路级噪声,其可能损坏SSD块中的现有数据,然后导致高读错误率和更长的读延迟。读刷新的方法通常用于通过周期性地将热读数据迁移到其他空闲块来避免读干扰错误,但是它对I/O(输入/输出)响应性产生相当大的负面影响。本文提出了写数据和读刷新操作的调度方法,以减轻读干扰带来的负面影响。具体来说,我们首先构建一个模型,通过参考块的P/E(Program/EQUIPMENT)周期和对块的累积读取计数的因素,根据估计的读取错误率将SSD块分为两类。然后,被密集读取的数据将被重定向到具有小读取错误率的块,因为即使数据将被大量请求,它对读取干扰也不敏感。此外,我们利用强化学习来预测两个I/O请求之间的空闲时间间隔,以便有目的地进行(部分)读取刷新操作。因此,它能够最大限度地减少对后续传入I/O请求的负面影响,并确保I/O响应能力。通过一系列的仿真测试几个现实的磁盘痕迹,我们表明,所提出的机制可以显着提高性能的指标的读错误率和I/O延迟。
Read disturb is a circuit-level noise in solid-state drives (SSDs), which may corrupt existing data in SSD blocks and then cause high read error rate and longer read latency. The approach of read refresh is commonly used to avoid read disturb errors by periodically migrating the hot read data to other free blocks, but it places considerable negative impacts on I/O (Input/Output) responsiveness. This article proposes scheduling approaches on write data and read refresh operations, to mitigate the negative effects caused by read disturb. To be specific, we first construct a model to classify SSD blocks into two categories according to the estimated read error rate by referring to the factors of block’s P/E (Program/Erase) cycle and the accumulated read count to the block. Then, the data being intensively read will be redirected to the block having a small read error rate, as it is not sensitive to read disturb even though the data will be heavily requested. Moreover, we take advantage of reinforcement learning to predict the idle interval between two I/O requests for purposely conducting (partial) read refresh operations. As a result, it is able to minimize negative impacts toward subsequent incoming I/O requests and to ensure I/O responsiveness. Through a series of emulation tests on several realistic disk traces, we demonstrate that the proposed mechanisms can noticeably yield performance improvements on the metrics of read error rate and I/O latency.
DCR:NAND 闪存存储系统的确定性崩溃恢复
DOI: 10.1109/tcad.2018.2878179
发表时间: 2019-12
影响因子: 2.9
作者:
Renhai Chen;Chi Zhang;Yi Wang;Zhaoyan Shen;Duo Liu;Zili Shao;Yong Guan
通讯作者: Yong Guan
DOI: 10.1145/1416944.1416949
发表时间: 2008-11
期刊: --
影响因子: --
作者:
D. Narayanan;Austin Donnelly;A. Rowstron
通讯作者: D. Narayanan;Austin Donnelly;A. Rowstron
DOI: 10.1016/0378-4754(87)90094-2
发表时间: 1986
期刊: --
影响因子: --
作者:
坂元 慶行;石黒 真木夫;北川 源四郎
通讯作者: 坂元 慶行;石黒 真木夫;北川 源四郎
DOI: --
发表时间: 2012-02
期刊: --
影响因子: --
作者:
Laura M. Grupp;John D. Davis;S. Swanson
通讯作者: Laura M. Grupp;John D. Davis;S. Swanson
DOI: 10.1109/irps.2017.7936387
发表时间: 2017-04
期刊: 2017 IEEE International Reliability Physics Symposium (IRPS)
影响因子: --
作者:
C. Zambelli;P. Olivo;L. Crippa;A. Marelli;R. Micheloni
通讯作者: C. Zambelli;P. Olivo;L. Crippa;A. Marelli;R. Micheloni