计算机科学与探索 ›› 2017, Vol. 11 ›› Issue (3): 341-354.DOI: 10.3778/j.issn.1673-9418.1604049

• 综述·探索 • 上一篇    下一篇

存储系统可靠性预测综述

李  静1,2,王  刚2,刘晓光2,李忠伟2+   

  1. 1. 中国民航大学 计算机科学与技术学院,天津 300300
    2. 南开大学 计算机与控制工程学院,天津 300350
  • 出版日期:2017-03-01 发布日期:2017-03-09

Review of Reliability Prediction for Storage System

LI Jing1,2, WANG Gang2, LIU Xiaoguang2, LI Zhongwei2+   

  1. 1. College of Computer Science and Technology, Civil Aviation University of China, Tianjin 300300, China
    2. College of Computer and Control Engineering, Nankai University, Tianjin 300350, China
  • Online:2017-03-01 Published:2017-03-09

摘要: 对存储系统的可靠性预测,可以用来评估、比较不同容错机制以及评价不同系统参数对存储系统可靠性的作用,有利于创建高可靠的存储系统。因此,存储系统可靠性预测的研究一直是领域热点之一。从硬盘单体和存储系统两种预测对象的角度,对近年来该领域的相关研究进行了介绍和分析。首先从硬盘个体和群体两种对象,主动容错和被动容错两种容错方式,以及纠删码和副本两种冗余机制几个角度分类介绍当前可靠性预测研究现状,然后进一步指明当前该领域研究中尚未解决的一些难题和未来可能的发展方向。分析表明,目前对副本存储系统和主动容错存储系统的可靠性预测研究尚显不足,是未来很有价值的研究方向。

关键词: 存储系统, 可靠性预测, 主动容错方式, 被动容错方式

Abstract: The reliability prediction for storage system, which is useful to assess trade-offs, compare schemes and estimate the effect of several parameters on storage system reliability, can help system designers and administrators to build storage systems with high reliability. So the research on reliability prediction is always one of the research    focuses in storage system. This paper makes careful introduction and analysis in the field of reliability prediction for storage system, respectively from two prediction objects—disk and storage system. Firstly, this paper carefully analyzes the current development status of storage system reliability prediction, from the perspectives of two objects—disk individual and disk family, two fault tolerant manners—proactive fault tolerant scheme and reactive fault tolerant scheme, and two redundant mechanisms—erasure code and replication. Then, this paper indicates the unresolved problems and the future trend in this field. From the analysis, this paper finds there are some weaknesses and drawbacks on the reliability prediction for replication storage and proactive fault tolerant systems, and they are the issues needing further study.

Key words: storage system, reliability prediction, proactive fault tolerant, reactive fault tolerant