海量数据存储系统的高可靠性关键技术研究与应用
|
摘要
随着信息技术的不断发展，数据日益成为人们日常生活中重要资源。据统计，2010年全球创建、存储和复制的数字信息总量已达到1.2ZB，2011年的数据量达到里程碑式的1.8ZB，而且这种增长还在加速，预计2015年将达到近8ZB。爆炸式增长的数据必然带来存储设备的持续增加。目前，海量数据存储环境下的现代数据中心的存储节点规模少则几万多则几十万，但在规模巨大的存储环境系统中，磁盘损毁或者存储节点失效已成为一种常态行为；与此同时，因网络连接设备或者存储节点其它元器件造成的数据不可访问或者丢失现象也时有发生。为了满足日益扩展的数据存储需求，人们对数据存储的可靠性，可用性等相关特性提出了更高的要求，传统的常规技术手段根本无法应对当前的形式，如何实现海量数据的低冗余度高可靠性存储已经成为业界面临的一个巨大挑战。
因而，本文针对构建低冗余度高可靠性海量数据存储系统的关键问题，在总结了当今数据可靠性增强理论和海量数据存储系统基本架构的基础上，对高性能数据容删数据布局算法以及高可靠性存储架构等方面进行了深入的研究，取得了一定的进展，具体如下：
1．针对已在数据存储系统中有广泛应用的RAID技术，提出了一种新的基于异或运算的水平阵列纠删码：EX-ENOD码。该码能够容许任意三列的随机删除错，并具有极大距离可分性质。根据该码的几何构造特点，提出了一种具有低计算复杂度的译码方法，该方法的计算复杂度要低于目前已知的其它可纠三列随机删除错纠删码的译码方法。同时，该编码方法具有明显的通用性，可以扩展应用到STAR码、EEOD码的译码过程。
2．针对海量存储系统规模不断扩大，可靠性要求不断提高的需求，本文将在{0,1}符号域上的范德蒙系统编码方法引入存储系统中来。该种编码方法继承了传统有限域上构建的范德蒙编码的参数不受存储节点规模、容错参数限制的优良特性，而且存储效率达到最高，同时该方法突破了传统有限域上构建的编码需要大量查询运算的弊端。基于该编码方法构建的存储系统，在保证其数据仍然可用的情况下，最多可容许系统内部一半的存储节点发生损毁，在该种情况下，系统仅需要与原数据相同的冗余数据量。
3．针对{0,1}符号域上编码矩阵的特点，根据编码矩阵中各行向量中“1”元素的分布，提出了可降低编译码计算复杂度的优化算法。文章同时针对传统译码重构过程重构带宽较高的缺点，提出了基于校验矩阵的译码方法，并根据校验矩阵列向量的特点和存储系统所需要重构数据的数目，给出了一种低带宽重构算法。该种低带宽重构算法，可以推广到所有构建在{0,1}符号域上的编码存储系统。
4．根据编码冗余策略数据布局的特点，设计了一种低冗余度高可靠性海量数据存储系统基础架构。系统将数据消冗和编码冗余可靠性增强技术纳入统一的基础架构，并针对编码冗余数据的分布特点，进行了存储节点的节能设计；针对数据使用特点提出了非均等存储及自适应读取策略；并提出了数据消冗与数据验证协同进行的运行策略。
With the development of information technology, data is increasingly becomingan important resource in our daily lives. According to the statistics, the total amount ofdata which the world created in2010, stored and replicated has reached1.2ZB, and in2011the amount is1.8ZB. It is expected to reach nearly8ZB in2015. The primaryproblem brought by the explosive growth of data is the increasing costs of storageequipment and management. Now, the scale of storage nodes in the modern data centeris ranging from tens of thousands to hundreds of thousands. Disk and storage nodefailures have become a normal behavior in the huge storage systems. At the same time,the network connection device and other components of storage node often becomeinvalid. In order to meet the expanding demand of reliability, availability, and otherrelevant characteristics of data storage, there need new techniques to cope with thecurrent situation. How to implement the mass data storage technology with lowredundancy and high reliability has become a huge challenge.
Thus, this dissertation focuses on the key technology of constructing the massdata storage system with low redundancy and high reliability. Basing on the longresearch of the theory of data reliability and the framework of the storage systems, wehave made some progress. The main contents of this dissertation are as follows:
1. As the RAID technology has already been widely used in the data storagesystem, we propose a new level array code based on the exclusive-OR operation:EX-ENOD code. This code can tolerate any random three failures and it has the beststorage efficiency. According to the geometrical configuration of the code, we proposea decoding method with low computational complexity. Further more thecomputational complexity of this method is much lower than those of other erasurecodes. Meanwhile, this decoding method can be applied to the STAR code, the EEODcode and so on.
2. In order to improve the reliability of the mass storage system, we design andimplement Vandermonde system encoding method that is based on {0,1} symboldomain in the mass storage system. This encoding method inherits the excellent features of Vandermonde code that is on the traditional finite field. There is no need toconsider the limits number of the storage nodes and fault-tolerant parameters. At thesame time, it can achieve the best storage efficiency. This method can avoid thedisadvantage of the traditional encoding on finite field which requires a lot of queryoperations. In particularly, this method is very suitable for the binary representation ofthe data in the computer system, so it is more suitable for the mass data storage thanothers.
3. According to the distribution of the "1" element in each row vector of theencoding matrix in {0,1} symbol domain, we propose an optimization algorithm whichcan reduce the computational complexity of encoding and decoding. At the same time,In order to reduce the high bandwidth that is needed in the traditional decodingreconstruction process, we propose a decoding method which is based on the paritycheck matrix. We also give a low-bandwidth reconstruction method according to thecharacteristics of the parity check matrix column vector. This low-bandwidthreconstruction algorithm can be extended to all the encoding storage systems that arebuilt on {0,1} symbol domain.
4. According to the characteristics of the data layout of erasure codes, we design amass storage system infrastructure with low redundancy and high reliability. Thissystem combine the data deduplication technology and encoding technology together.Basing on the structural characteristics of the erasure code, we extend an energy-savingand self-adaption access strategy to the storage system. At the same time, we alsointegrate the data redundant elimination and data validation together in the system.
因而，本文针对构建低冗余度高可靠性海量数据存储系统的关键问题，在总结了当今数据可靠性增强理论和海量数据存储系统基本架构的基础上，对高性能数据容删数据布局算法以及高可靠性存储架构等方面进行了深入的研究，取得了一定的进展，具体如下：
1．针对已在数据存储系统中有广泛应用的RAID技术，提出了一种新的基于异或运算的水平阵列纠删码：EX-ENOD码。该码能够容许任意三列的随机删除错，并具有极大距离可分性质。根据该码的几何构造特点，提出了一种具有低计算复杂度的译码方法，该方法的计算复杂度要低于目前已知的其它可纠三列随机删除错纠删码的译码方法。同时，该编码方法具有明显的通用性，可以扩展应用到STAR码、EEOD码的译码过程。
2．针对海量存储系统规模不断扩大，可靠性要求不断提高的需求，本文将在{0,1}符号域上的范德蒙系统编码方法引入存储系统中来。该种编码方法继承了传统有限域上构建的范德蒙编码的参数不受存储节点规模、容错参数限制的优良特性，而且存储效率达到最高，同时该方法突破了传统有限域上构建的编码需要大量查询运算的弊端。基于该编码方法构建的存储系统，在保证其数据仍然可用的情况下，最多可容许系统内部一半的存储节点发生损毁，在该种情况下，系统仅需要与原数据相同的冗余数据量。
3．针对{0,1}符号域上编码矩阵的特点，根据编码矩阵中各行向量中“1”元素的分布，提出了可降低编译码计算复杂度的优化算法。文章同时针对传统译码重构过程重构带宽较高的缺点，提出了基于校验矩阵的译码方法，并根据校验矩阵列向量的特点和存储系统所需要重构数据的数目，给出了一种低带宽重构算法。该种低带宽重构算法，可以推广到所有构建在{0,1}符号域上的编码存储系统。
4．根据编码冗余策略数据布局的特点，设计了一种低冗余度高可靠性海量数据存储系统基础架构。系统将数据消冗和编码冗余可靠性增强技术纳入统一的基础架构，并针对编码冗余数据的分布特点，进行了存储节点的节能设计；针对数据使用特点提出了非均等存储及自适应读取策略；并提出了数据消冗与数据验证协同进行的运行策略。
With the development of information technology, data is increasingly becomingan important resource in our daily lives. According to the statistics, the total amount ofdata which the world created in2010, stored and replicated has reached1.2ZB, and in2011the amount is1.8ZB. It is expected to reach nearly8ZB in2015. The primaryproblem brought by the explosive growth of data is the increasing costs of storageequipment and management. Now, the scale of storage nodes in the modern data centeris ranging from tens of thousands to hundreds of thousands. Disk and storage nodefailures have become a normal behavior in the huge storage systems. At the same time,the network connection device and other components of storage node often becomeinvalid. In order to meet the expanding demand of reliability, availability, and otherrelevant characteristics of data storage, there need new techniques to cope with thecurrent situation. How to implement the mass data storage technology with lowredundancy and high reliability has become a huge challenge.
Thus, this dissertation focuses on the key technology of constructing the massdata storage system with low redundancy and high reliability. Basing on the longresearch of the theory of data reliability and the framework of the storage systems, wehave made some progress. The main contents of this dissertation are as follows:
1. As the RAID technology has already been widely used in the data storagesystem, we propose a new level array code based on the exclusive-OR operation:EX-ENOD code. This code can tolerate any random three failures and it has the beststorage efficiency. According to the geometrical configuration of the code, we proposea decoding method with low computational complexity. Further more thecomputational complexity of this method is much lower than those of other erasurecodes. Meanwhile, this decoding method can be applied to the STAR code, the EEODcode and so on.
2. In order to improve the reliability of the mass storage system, we design andimplement Vandermonde system encoding method that is based on {0,1} symboldomain in the mass storage system. This encoding method inherits the excellent features of Vandermonde code that is on the traditional finite field. There is no need toconsider the limits number of the storage nodes and fault-tolerant parameters. At thesame time, it can achieve the best storage efficiency. This method can avoid thedisadvantage of the traditional encoding on finite field which requires a lot of queryoperations. In particularly, this method is very suitable for the binary representation ofthe data in the computer system, so it is more suitable for the mass data storage thanothers.
3. According to the distribution of the "1" element in each row vector of theencoding matrix in {0,1} symbol domain, we propose an optimization algorithm whichcan reduce the computational complexity of encoding and decoding. At the same time,In order to reduce the high bandwidth that is needed in the traditional decodingreconstruction process, we propose a decoding method which is based on the paritycheck matrix. We also give a low-bandwidth reconstruction method according to thecharacteristics of the parity check matrix column vector. This low-bandwidthreconstruction algorithm can be extended to all the encoding storage systems that arebuilt on {0,1} symbol domain.
4. According to the characteristics of the data layout of erasure codes, we design amass storage system infrastructure with low redundancy and high reliability. Thissystem combine the data deduplication technology and encoding technology together.Basing on the structural characteristics of the erasure code, we extend an energy-savingand self-adaption access strategy to the storage system. At the same time, we alsointegrate the data redundant elimination and data validation together in the system.
引文
[1] J. Gantz, D. Reinsel. The digital universe decade–are you ready?[R]. USA: Internationla DataCorporation, July16,2010
[2] J. Gantz. The diverse and exploding digital universe: An updated forecast of worldwideinformation growth through2011[R]. USA: Internationla Data Corporation (IDC), July20,2008
[3]罗英伟,汪小林,尹冬生,等.信息存储与管理[M].北京:人民邮电出版社,2010,48-58
[4] J. Hennessy, D. Patterson.计算机系统结构——量化研究方法（第四版）[M].北京:电子工业出版社,2010,248-297
[5]王意洁,孙伟东,周松,等.云计算环境下的分布式存储关键技术[J].软件学报,2012,23(4):962-986
[6] S. Ghemawat, H. Gobioff, S. Leung. The Google file system[C].9th ACM Symposium onOperating Systems Principles Proceedings, USA,2003,29-43
[7] B. Calder, J. Wang, A. Ogus, et al. Windows Azure storage: A highly available cloud storageservice with strong consistency[C]. Twenty-Third ACM Symposium on Operating SystemsPrinciples (SOSP2011) Proceedings, New York,2011,143-157
[8] G. DeCandia, D. Hastorun, M. Jampani, et al. Dynamo: Amazon’s highly available key-valuestore[C]. Twenty-First ACM SIGOPS Symposium on Operating Systems PrinciplesProceedings, New York,2007,205-220
[9] K. Shvachko, H. Kuang, S. Radia, et al. The Hadoop distributed file system[C]. IEEETwenty-Sixth Symposium on Massive Storage Systems and Technologies Proceedings,Washington.DC,2010,1-10
[10] A.Lakshman, P.Malik, Cassandra: a decentralized structured storage system [J], SIGOPS Oper.Syst. Rev.2010,44(2):35-40
[11] E.Pinheiro, W.D.Weber, L.A.Barroso. Failure trends in a large disk drive population[C].Proceedings of the5th USENIX conference on File and Storage Technologies, San Jose,2007,17-28
[12] B.Schroeder, G.A.Gibson. Disk failures in the real world: what does an MTTF of1,000,000hours mean to you?[C]. Proceedings of the5th USENIX conference on File and StorageTechnologies, San Jose,2007,1-16
[13] L.N.Bairavasun, G.R.Goodson, S.Pasupathy, et al. An analysis of latent sector errors in diskdrives[C]. Proceedings of the2007ACM SIGMETRICS International Conference onMeasurement and Modeling of Computer Systems, San Diego,2007,289-300
[14] R.T.Kaushik, M.Bhandarkar. Greenhdfs: towards an energy-conserving, storage efficient,hybrid hadoop compute cluster[C]. In Proceedings of the2010international conference onPower aware computing and systems, Berkeley,2010,1-9
[15] N.Joukov, J.Sipek. Greenfs: making enterprise computers greener by protecting them better [J].SIGOPS Oper. Syst. Rev.2008,42(3):69-80
[16] M.Blaum, J.Brady, J. Bruck, et al. EVENODD: An efficient scheme for tolerating double diskfailures in RAID architectures [J]. IEEE Transactions on Computers,1995,45(2):192-202
[17] M.Blaum, J.Bruck, A.Vardy, et al. MDS array codes with independent parity symbols[J], IEEETransactions on Information Theory,1996,42(3):529-542
[18] M.Blaum, J.Brady, J.Bruck, et al. The EVENODD code and its generalization [M]. HighPerformance Mass Storage and Parallel I/O: Technologies and Applications, New York:JohnWiley,2001,187-208
[19] B. Johannes, K. Malik, K. Richard. An XOR-Based erasure-resilient coding schme[R].California: International Computer Science Institute,1995,11-21
[20] M. Blaum. A Family of MDS array codes with minimal number of encodingoperations[C].IEEE International Symposium on Information Theory, USA,2006,2784-2788
[21] L. Xu, J. Bruck. X-code: MDS array codes with optimal encoding [J]. IEEE Transactions onInformation Theory,1999,45(1):272-276
[22] L. Xu, V. Bohossian, J. Bruck, et al. Low density MDS codes and factors of complete graphs[J]. IEEE Transaction on Information Theory,1999,45(6):1817-1826
[23] Y. Cassuto, J. Bruck. Cyclic lowest density MDS array codes [J]. IEEE Transaction onInformation Theory,2009,55(4):1721-1729
[24] Z. Wang, A. G. Dimakis, J. Bruck. Rebuilding for array codes in distributed storagesystems[C]. Workshop on the Application of Communication Theory to Emerging MemoryTechnologies Proceedings, New York,2010,1-7
[25] L. Xu, J. Bruck. Highly available distributed storage systems [J]. Lecture Notes in Control andInformation Sciences,1999,21(2):307-330
[26] C. Huang, L. Xu. STAR: An efficient coding scheme for correcting triple storage node failures[J]. IEEE Transactions on Computers,2008,57(7):889-901
[27] J. Plank, A. Buchsbaum, T. Bradley, et al. Minimum density RAID-6codes [J]. ACMTransactions on Storage,2011,6(4):16-37
[28] J. Plank. The RAID-6liberation codes[C]. In6thUSENIX Conference on File and StorageTechnologies Proceedings, Berkeley,2008,97-110
[29] J.S.Plank, J.Luo, C.D Schuman, et al. A performance evaluation and examination ofopen-Source erasure coding libraries for storage[C]. In the7th Usenix Conference on File andStorage Technologies, San Francisco,2009,336-345
[30] G.Michael. J.S.Thomason. A practical analysis of low-density parity-check erasure codes forwide-area storage applications[C]. In International Conference on Dependable Systems andNetworks, USA,2004,201-224
[31] O.Khan, R.Burns, J.S.Plank, et al. Rethinking erasure codes for cloud file systems: minimizingI/O for recovery and degraded reads[C]. In Proc. of USENIX FAST, San Jose,2012,213-224
[32]万武南,吴震,陈运,等.一种基于3容错阵列码的RAID数据布局[J].计算机学报,2007,30(10):1721-1730
[33]万武南.分布式安全存储系统纠删码技术的研究[D].北京:中国科学院研究生院,2006,97-105
[34]万武南,索望,张金全,等.基于EEOD码的一种有效的数据分布策略[J].电子科技大学学报,2007,36(3):834-837
[35]孟庆春.应用于分布式存储系统上的纠删码技术研究[D].北京:中国科学院研究生院,2007,57-96
[36]陈峥.一类新的阵列纠删码理论及应用研究[D].北京:中国科学院研究生院,2009,43-56
[37]张永成.一种高性能数据容灾系统的关键技术研究[D].北京:中国科学院研究生院,2009,97-105
[38] Z.Chen, X.J.Wang, S.Cao. Precoding method in raptor code based on B-J code[C].International Conference on Broadband Network&Multimedia Technology, Dalian,2007,121-124
[39] Z.Chen, X.J.Wang, Y.L.Jin, et al. Exploring fault-tolerant distributed storage system using GEcode[C]. The Fifth International Conference on Embedded Software and Systems, Beijing,2008,142-148
[40] A.Thomasian. Multi-level RAID for very large disk arrays [J]. SIGMETRICS PerformanceEvaluation Review,2006,33(4):17-22
[41] D.A.Patterson, G.Gibson, R.H.Katz. A case for redundant arrays of inexpensive disks[C].ACM International Conference on Management of Data Proceedings, New York,1988,109-116
[42] S. Rhea, C. Wells, P. Eaton, et al. Maintenance-free global data storage [J]. IEEE InternetComputing,2001,5(5):40-49
[43] B.Moore. ZFS:The last word in file systems [EB/OL] http://www.sun.com/software/solarisJuly17,2004
[44] P.H.Carns, W.B.Ligon, R.B. Ross, et al. PVFS: A parallel file system for Linux clusters [C]. Inthe4th Annual Linux Showcase and Conference, California,2000,317-327
[45] W. Tantisiriroj, S. Patil, G. Gibson. Data-intensive file systems for Internet services: A rose byany other name[R]. Pennsylvania:Carnegie Mellon University,2008,11-21
[46] B.Welch, M.Unangst, Z.Abbasi, etal. Scalable performance of the panasas parallel filesystem[C]. Proceedings of the6th USENIX Conference on File and Storage Technologies,California，2008,223-246
[47] K.Shvachko, H.Kuang, S.Radia, et al. The Hadoop distributed file system[C]. IEEETwenty-Sixth Symposium on Massive Storage Systems and Technologies Proceedings,Washington,2010,1-10
[48] R.T.Kaushik, L.Cherkasova, R.Campbell, et al. Lightning: self-adaptive, energy-conserving,multi-zoned, commodity green cloud storage system[C]. In Proceedings of the19th ACMInternational Symposium on High Performance Distributed Computing, New York,2010,332-335
[49] Y. Saito. FAB: Building distributed enterprise disk arrays from commodity components[C].International Conference Architectural Support for Programming Languages and OperatingSystems, Boston,2004,48-58
[50] A.E.Malek. UrsaMinor: Versatile cluster-based storage[C]. USENIX Conference on File andStorage Technologies, Pennsylvania,2005,59-72
[51] E. Anderson, X. Li, A. Merchant, et al. Efficient eventual eventual consistency in Pahoehoe,an erasure-coded key-blob archive[C]. International Conference on Dependable Systems andNetworks, Chicago,2010,181-190
[52] M.K.Aguilera, R.Janakiraman, L.Xu. Using Erasure codes efficiently for storage in adistributed system[C] In International Conference on Dependable Systems and Networks,Boston,2005,234-247
[53] A. Thusoo, D. Borthakur, R. Murthy, et al. Data warehousing and analytics infrastructure atFacebook[C]. In Proceedings of the18th ACM SIGMOD Symposium2010, Indianaplis,2010,432-443
[54] R. Buyya, T. Corte, H. Jin. Petal: Distributed virtual disks[C]. In Proceeding of the HighPerformance Mass Storage and Parallel I/O: Technologies and Applications, Piscataway,2002,420-430
[55] F. Bek, M. F. Kaashoek, D. Karger, et al. Wide-Area cooperative storage with CFS[C]. InProceeding of the18th ACM Symp On Operating System Principles, Banff,2001,202-215
[56] J. A. Andy. A generalized replica placement strategy to optimize latency in a wide areadistributed storage system[C]. Proceedings of the2008International Workshop on Data-awareDistributed Computing, New York,2008,49-54
[57] Y. Ding, Y. Lu. Automatic data placement and replication in grids[C]. In Proc of the HIPC.Kochi,2009,30-39
[58] A. Adya, W. Bolosky, M. Castro, et al. Farsite: Federated, available, and reliable storage for anincompletely trusted environment[C]. Proceedings of the7th Symposium on OperatingSystems Design and Implementation, San Diego,2002,435-446
[59] R.Honicky, E. Miller. Replication under scalable hashing: A family of algorithms for scalabledecentralized data distribution[C]. Proceedings of the18th International Parallel&Distributed Processing, London,2004,216-231
[60] S. Weil, S. Brandt, E. Miller, et al. Grid resource management-crush: controlled, scalable,decentralized placement of replicated data [C]. Proceedings of the18th ACM Symposium OnOperating System Principles, USA,2006,221
[61] K. Shi, A. S.Thanamani. A new replica creation and placement algorithm for data gridenvironment[C]. Proceedings of the2010on Data Storage and Data Engineering, Bangalore,2010,265-269
[62] Q. Chen, B. Y. Zhang. Dynamic replica placement and location strategies for data grid[C].Proceedings of the2008on Computer Science and Software Engineering, Wuhan,2008,35-40
[63] A. S. Weil, A. B. Scott, L. Miller. Ceph: A scalable, high-performance distributed filesystem[C]. Proceedings of the7th Symposium on Operating Systems Design andImplementation, Santa Cruz,2006,307-320
[64] J. Baker, C. Bond, J. Corbett, et al. Megastore: providing scalable, highly available storage forinteractive services. Conference on Innovative Data Systems Research, Asilomar,2011,106-117
[65] M. Li, J. Su, W. Zheng. Grid Codes: Strip-based erasure codes with high fault tolerance forstorage systems [J]. ACM Transactions on Storage,2009,4(4):15:1-15:22
[66] G. Harihara，B. Janakiram，M. G. Chandra，et al. SpreadStore: A LDPC erasure code schemefor distributed storage system[C]. Proceedings of the2010International Conference on DataStorage and Data Engineering, Bangalore,2010,87-93
[67] M. K. Greenan，L. X. Zhou，J. J. Wylie. Flat XOR-based erasure codes in storage systemsConstructions, efficient recovery, and tradeoffs[C]. IEEE26th Symposium on Mass StorageSystems and Technologies, Incline Village,2010,1-14
[68] S. Reed, G. Solomon. Polynomial codes over certain finite fields [J]. Journal of the Society forIndustrial and Applied Mathematics,1960,8(2):300-304
[69] M. Rabin. Efficient dispersal of information for security loading balancing and fault tolerance[J]. Journal of the Association for Computing Machinery,1989,36(2):335-348
[70] Y. Chen, J. Edler, A. Goldberg, et al. A prototype implementation of archival intermemory[C].Fourth ACM Conference on Digital libraries Proceedings, New York,1999,28-37
[71] J. Plank. A tutorial on Reed-Solomon coding for fault-tolerance in RAID-like systems [J].Software-Practice and Experience,1997,27(9):995-1012
[72] J. Plank, Y. Ding. Note: Correction to the1997tutorial on Reed-Solomon coding [J].Software-Practice and Experience,2005,35(2):189-194
[73] R. M. Roth, A. Lempel. On MDS codes via Cauchy Matrices [J]. IEEE Transaction onInformation Theory,1989,35(6),1314-1319
[74] J. Blomer, M. Kalfane, M. Karpinski, et al. An XOR-based erasure-resilient coding scheme[R].Berkeley:International Computer Science Institute,1995,11-23
[75] J. Plank, L. Xu. Optimizing Cauchy Reed-Solomon codes for fault-tolerant network storageapplications[C]. IEEE International Symposium on Network Computing and ApplicationsProceedings, Cambridge,2006,173-180
[76] J. Plank, C. Schuman, B. Robison. Heuristics for optimizing matrix-based erasure codes forfault-tolerant storage systems[C]. International Conference on Dependable Systems andNetworks Proceedings, Boston,2012,1-12
[77] A. Dimakis, P. Godfrey, J. Wainwright, et al. Network coding for distributed storage systems
[C].26th IEEE International Conference on Computer Communications Proceedings, Alaska,2007,200-208
[78] A. Dimakis, P. Godfrey, Y. Wu, et al. Network coding for distributed storage systems [J]. IEEETransactions on Information Theory,2010,56(9):4539-4551
[79] A. Dimakis, K. Ramchandran, Y. Wu, et al. A survey on network codes for distributed storage
[C]. Proceedings of the IEEE, Carolina,2011,476-489
[80]常乾,许胤龙,项利萍,等.基于EVENODD码的单盘故障快速恢复算法[J].计算机应用与软件,2011,28(6):15-18
[81] L. Xiang, Y. Xu, C. S. Lui, et al. Optimal recovery of single disk failure in RDP code storagesystems [J]. ACM SIGMETRICS Performance Evaluation Review,2010,38(1):119-130
[82] K. V. Rashmi, N. B. Shah, P. V. Kumar, et al. Explicit construction of optimal exactregenerating codes for distributed storage [J]. Communication, Control,and Computing,2009,22(6):211-243
[83] Y. Hu, Y. Xu, X. Wang, et al. Cooperative recovery of distributed storage system frommultiple losses with network coding[J]. IEEE Journal on Selected Areas in Communication,2010,28(2):268-276
[84] L. You, K. Pollack, D. Long, Deep store: An archival storage system architecture[C].Proceedings of the21st IEEE International Conference on Data Engineering, California,2005,804-815
[85] S.Quinlan，S.Dorward, Venti: A new approach to archival storage[C]. In Proceedings of the2002Conference on File and Storage Technologies (FAST), California,2002,89-101
[86] M. Lillibridge, K. Eshghi, D. Bhagwat, et al. Sparse indexing: Large scale, inlinededuplication using sampling and locality [R]. Proceeding of the7thUSENIX Conference onFile and Storage Technologies, San Francisco,2009,111-123
[87] A. Mut, C. Benjie, D. Mazières, et al. A low bandwidth network file system [C] Proceeding ofthe18thACM Symposium on Principles, New York,2001,174-187
[88] M. O. Rabin. Fingerprinting by random polynomials[R]. Boston: Centerfor Research inComputing Technology, Harvard University,1981,23-34
[89] C. Y. Liu, Y. P. Lu, D David, etal. ADMAD: Application-driven metadata aware de-duplicationarchival storage system [C]. Proceeding of the25thIEEE Conf Mass Storage Syst and Technol.Piscataway,2008,29-35
[90]敖莉，舒继武，李明强.重复数据删除技术[J]，计算机研究与发展，2010,21(5):916-929
[91] W. E. Binary. HP storage works D2D4000backup system: a report and full Performance teston Hewlett-Packard's SME data deduplication appliance[R]. California: Hewlett-PackardDevelopment Company, April,2008
[92] L.Aronovich, R.Asher, E.Bachmat, et al. The design of a similarity based deduplicationsystem[C]. In Proceedings of SYSTOR2009:The Israeli Experimental Systems Conference，Israeli,2009,11-26
[93] EMC Corporation. EMC centera: content addressed storage system[R]. Massachusetts: EMCCorporation, May,2002
[94] H. Li, H. Patterson. Avoiding the disk bottleneck in the data domain deduplication filesystem[C]. Proceedings of the6th USENIX Conference on File And Storage Technologies,California,2008,1-14
[95] A. Adya, W. J. Bolosky, M. Castro, et al. FARSITE: Federated, available, and reliable storagefor an incompletely trusted environment[C]. Proceedings of the5th Symposium On OperatingSystems Design and Implementation (OSDI2002), New York,USA: ACM Press,2002,1-14
[96] W. J. Losky, S. Corbin, D. Goebel, et al. Single instance storage in Windows2000[C].Proceedings of the4th Usenix Windows System, Berkeley,2000,13-24
[97] P. Sehgal, V. Tarasov, E. Zadok. Optimizing energy and performance for server-class filesystem workloads [J]. IEEE Trans on Storage,2010,10(6):10-31
[98] J. Kim, D. Rotem. Energy proportionality for disk storage using replication[C]. In Proceedingsof the14th International Conference on Extending Database Technology, New York,2011:81-92
[99] E. Otoo, D. Rotem, S. C. Tsao. Analysis of trade-off between power saving and response timein disk storage systems[C]. In Proceedings of IEEE International Symposium on Parallel andDistributed Processing, Berkeley,2009,1012-1031
[100] H. Huang, W. D. Huang, G. K. Shin. FS2: Dynamic data replication in free disk space forimproving disk performance and energy Consumption [C] Proceedings of the20th ACMSymposium on Operating Systems Principles, Brighton,2005:263-276
[101] S. Gurumurthi, A. Sivasub, K. M. Demir, et al. DRPM:Dynamic speed control for powermanagement in server class disks [C] The ACM Proceedings of the International Symposiumon Computer Architecture, New York,2003,169-181
[102] S. Sankar, S. Gurumurthi, M. R.Stan. Intra-disk parallelism: An idea whose time has come [R].Virginia: University of Virginia, February,2008
[103] T. Bisson, S. Brandt, D. Long. A hybrid disk-aware spin-down algorithm with I/O subsystemsupport[C]. In Proceedings of the26th IEEE International Performance Computing andCommunications Conference, Los Alamitos,2007,236-245
[104] H. J. Lee, K. H. Lee, S. H. Noh. Augmenting raid with an ssd for energy relief[C]. InProceedings of the2008Conference on Power Aware Computing and Systems, Berkeley,2008,12-18
[105] D. Narayanan, E. Thereska, M. Donnelly, et al. Migrating server storage to ssds: analysis oftrades[C]. Proceedings of the4th ACM European Conference on Computer systems, NewYork,2009,145-158
[106] Y. Deng. What is the future of disk drives, death or rebirth?[J]. ACM Computing Surveys,2011,42(23):123-27
[107] E. Thereska, A. Donnelly, D. Narayanan. Sierra: practical power-proportionality for datacenter storage[C]. In Proceedings of the Sixth Conference on Computer Systems, New York,2006,169-182
[108]万哲先.代数和编码[M].北京:高等教育出版社,2007,201-234
[109]冯克勤.纠错码的代数理论[M].北京:清华大学出版社,2005,14-22
[110] S. Roman. Coding and information theory [M]. New York: Springer-verlag,1992,254-263
[111] J. S. Plank. XORs Lower bounds and MDS codes for storage[R]. Tennessee: University ofTennessee, May,2010
[112]董欢庆,李战怀,林伟. RAID-VCR:一种能够承受三个磁盘故障的RAID结构[J].计算机学报,2006,29(3):792-800
[113]魏青松.大规模分布式存储技术研究[D].成都：电子科技大学，2004
[114]周旭.面向Internet的大规模分布式存储技术研究[D].成都：电子科技大学，2004
[115]吴素贞.磁盘阵列高可用技术研究[D].武汉：华中科技大学，2010
[116]李洁琼.海量存储资源管理关键技术研究[D].武汉：华中科技大学，2010
[117]刘刚.分布式存储系统中基于纠错码的容错机制研究[D].武汉：华中科技大学，2007
[118] M. Li, J. Shu. Preventing silent data corruptions from propagating during data reconstruction[J]. IEEE Transactions on Computers,2010,59(12):1611-1624.
[119] M. Li, J. Shu. DACO: High-performance disk architecture designed especially for large scaleerasure-coded storage systems [J]. IEEE Transactions on Computers,2010,59(10):1350-1362.
[120] M. Li, J. Shu. On the equivalence between the B-Code constructions and perfectone-factorizations[C]. Proceedings of the2010IEEE International Symposium on InformationTheory (ISIT’10), Austin,2010,993-996.
[121] M. Li, J. Shu. On cyclic lowest density MDS array codes constructed using starters.Proceedings of the2010IEEE International Symposium on Information Theory (ISIT’10),Austin,2010,1315-1319.
[122] P. Corbett, B. English, A.Goel, et al. Row diagonal parity for double disk failure correction[C].3rdUsenix Conference on File and Storage Technologies, San Francisco,2004,121-132
[2] J. Gantz. The diverse and exploding digital universe: An updated forecast of worldwideinformation growth through2011[R]. USA: Internationla Data Corporation (IDC), July20,2008
[3]罗英伟,汪小林,尹冬生,等.信息存储与管理[M].北京:人民邮电出版社,2010,48-58
[4] J. Hennessy, D. Patterson.计算机系统结构——量化研究方法（第四版）[M].北京:电子工业出版社,2010,248-297
[5]王意洁,孙伟东,周松,等.云计算环境下的分布式存储关键技术[J].软件学报,2012,23(4):962-986
[6] S. Ghemawat, H. Gobioff, S. Leung. The Google file system[C].9th ACM Symposium onOperating Systems Principles Proceedings, USA,2003,29-43
[7] B. Calder, J. Wang, A. Ogus, et al. Windows Azure storage: A highly available cloud storageservice with strong consistency[C]. Twenty-Third ACM Symposium on Operating SystemsPrinciples (SOSP2011) Proceedings, New York,2011,143-157
[8] G. DeCandia, D. Hastorun, M. Jampani, et al. Dynamo: Amazon’s highly available key-valuestore[C]. Twenty-First ACM SIGOPS Symposium on Operating Systems PrinciplesProceedings, New York,2007,205-220
[9] K. Shvachko, H. Kuang, S. Radia, et al. The Hadoop distributed file system[C]. IEEETwenty-Sixth Symposium on Massive Storage Systems and Technologies Proceedings,Washington.DC,2010,1-10
[10] A.Lakshman, P.Malik, Cassandra: a decentralized structured storage system [J], SIGOPS Oper.Syst. Rev.2010,44(2):35-40
[11] E.Pinheiro, W.D.Weber, L.A.Barroso. Failure trends in a large disk drive population[C].Proceedings of the5th USENIX conference on File and Storage Technologies, San Jose,2007,17-28
[12] B.Schroeder, G.A.Gibson. Disk failures in the real world: what does an MTTF of1,000,000hours mean to you?[C]. Proceedings of the5th USENIX conference on File and StorageTechnologies, San Jose,2007,1-16
[13] L.N.Bairavasun, G.R.Goodson, S.Pasupathy, et al. An analysis of latent sector errors in diskdrives[C]. Proceedings of the2007ACM SIGMETRICS International Conference onMeasurement and Modeling of Computer Systems, San Diego,2007,289-300
[14] R.T.Kaushik, M.Bhandarkar. Greenhdfs: towards an energy-conserving, storage efficient,hybrid hadoop compute cluster[C]. In Proceedings of the2010international conference onPower aware computing and systems, Berkeley,2010,1-9
[15] N.Joukov, J.Sipek. Greenfs: making enterprise computers greener by protecting them better [J].SIGOPS Oper. Syst. Rev.2008,42(3):69-80
[16] M.Blaum, J.Brady, J. Bruck, et al. EVENODD: An efficient scheme for tolerating double diskfailures in RAID architectures [J]. IEEE Transactions on Computers,1995,45(2):192-202
[17] M.Blaum, J.Bruck, A.Vardy, et al. MDS array codes with independent parity symbols[J], IEEETransactions on Information Theory,1996,42(3):529-542
[18] M.Blaum, J.Brady, J.Bruck, et al. The EVENODD code and its generalization [M]. HighPerformance Mass Storage and Parallel I/O: Technologies and Applications, New York:JohnWiley,2001,187-208
[19] B. Johannes, K. Malik, K. Richard. An XOR-Based erasure-resilient coding schme[R].California: International Computer Science Institute,1995,11-21
[20] M. Blaum. A Family of MDS array codes with minimal number of encodingoperations[C].IEEE International Symposium on Information Theory, USA,2006,2784-2788
[21] L. Xu, J. Bruck. X-code: MDS array codes with optimal encoding [J]. IEEE Transactions onInformation Theory,1999,45(1):272-276
[22] L. Xu, V. Bohossian, J. Bruck, et al. Low density MDS codes and factors of complete graphs[J]. IEEE Transaction on Information Theory,1999,45(6):1817-1826
[23] Y. Cassuto, J. Bruck. Cyclic lowest density MDS array codes [J]. IEEE Transaction onInformation Theory,2009,55(4):1721-1729
[24] Z. Wang, A. G. Dimakis, J. Bruck. Rebuilding for array codes in distributed storagesystems[C]. Workshop on the Application of Communication Theory to Emerging MemoryTechnologies Proceedings, New York,2010,1-7
[25] L. Xu, J. Bruck. Highly available distributed storage systems [J]. Lecture Notes in Control andInformation Sciences,1999,21(2):307-330
[26] C. Huang, L. Xu. STAR: An efficient coding scheme for correcting triple storage node failures[J]. IEEE Transactions on Computers,2008,57(7):889-901
[27] J. Plank, A. Buchsbaum, T. Bradley, et al. Minimum density RAID-6codes [J]. ACMTransactions on Storage,2011,6(4):16-37
[28] J. Plank. The RAID-6liberation codes[C]. In6thUSENIX Conference on File and StorageTechnologies Proceedings, Berkeley,2008,97-110
[29] J.S.Plank, J.Luo, C.D Schuman, et al. A performance evaluation and examination ofopen-Source erasure coding libraries for storage[C]. In the7th Usenix Conference on File andStorage Technologies, San Francisco,2009,336-345
[30] G.Michael. J.S.Thomason. A practical analysis of low-density parity-check erasure codes forwide-area storage applications[C]. In International Conference on Dependable Systems andNetworks, USA,2004,201-224
[31] O.Khan, R.Burns, J.S.Plank, et al. Rethinking erasure codes for cloud file systems: minimizingI/O for recovery and degraded reads[C]. In Proc. of USENIX FAST, San Jose,2012,213-224
[32]万武南,吴震,陈运,等.一种基于3容错阵列码的RAID数据布局[J].计算机学报,2007,30(10):1721-1730
[33]万武南.分布式安全存储系统纠删码技术的研究[D].北京:中国科学院研究生院,2006,97-105
[34]万武南,索望,张金全,等.基于EEOD码的一种有效的数据分布策略[J].电子科技大学学报,2007,36(3):834-837
[35]孟庆春.应用于分布式存储系统上的纠删码技术研究[D].北京:中国科学院研究生院,2007,57-96
[36]陈峥.一类新的阵列纠删码理论及应用研究[D].北京:中国科学院研究生院,2009,43-56
[37]张永成.一种高性能数据容灾系统的关键技术研究[D].北京:中国科学院研究生院,2009,97-105
[38] Z.Chen, X.J.Wang, S.Cao. Precoding method in raptor code based on B-J code[C].International Conference on Broadband Network&Multimedia Technology, Dalian,2007,121-124
[39] Z.Chen, X.J.Wang, Y.L.Jin, et al. Exploring fault-tolerant distributed storage system using GEcode[C]. The Fifth International Conference on Embedded Software and Systems, Beijing,2008,142-148
[40] A.Thomasian. Multi-level RAID for very large disk arrays [J]. SIGMETRICS PerformanceEvaluation Review,2006,33(4):17-22
[41] D.A.Patterson, G.Gibson, R.H.Katz. A case for redundant arrays of inexpensive disks[C].ACM International Conference on Management of Data Proceedings, New York,1988,109-116
[42] S. Rhea, C. Wells, P. Eaton, et al. Maintenance-free global data storage [J]. IEEE InternetComputing,2001,5(5):40-49
[43] B.Moore. ZFS:The last word in file systems [EB/OL] http://www.sun.com/software/solarisJuly17,2004
[44] P.H.Carns, W.B.Ligon, R.B. Ross, et al. PVFS: A parallel file system for Linux clusters [C]. Inthe4th Annual Linux Showcase and Conference, California,2000,317-327
[45] W. Tantisiriroj, S. Patil, G. Gibson. Data-intensive file systems for Internet services: A rose byany other name[R]. Pennsylvania:Carnegie Mellon University,2008,11-21
[46] B.Welch, M.Unangst, Z.Abbasi, etal. Scalable performance of the panasas parallel filesystem[C]. Proceedings of the6th USENIX Conference on File and Storage Technologies,California，2008,223-246
[47] K.Shvachko, H.Kuang, S.Radia, et al. The Hadoop distributed file system[C]. IEEETwenty-Sixth Symposium on Massive Storage Systems and Technologies Proceedings,Washington,2010,1-10
[48] R.T.Kaushik, L.Cherkasova, R.Campbell, et al. Lightning: self-adaptive, energy-conserving,multi-zoned, commodity green cloud storage system[C]. In Proceedings of the19th ACMInternational Symposium on High Performance Distributed Computing, New York,2010,332-335
[49] Y. Saito. FAB: Building distributed enterprise disk arrays from commodity components[C].International Conference Architectural Support for Programming Languages and OperatingSystems, Boston,2004,48-58
[50] A.E.Malek. UrsaMinor: Versatile cluster-based storage[C]. USENIX Conference on File andStorage Technologies, Pennsylvania,2005,59-72
[51] E. Anderson, X. Li, A. Merchant, et al. Efficient eventual eventual consistency in Pahoehoe,an erasure-coded key-blob archive[C]. International Conference on Dependable Systems andNetworks, Chicago,2010,181-190
[52] M.K.Aguilera, R.Janakiraman, L.Xu. Using Erasure codes efficiently for storage in adistributed system[C] In International Conference on Dependable Systems and Networks,Boston,2005,234-247
[53] A. Thusoo, D. Borthakur, R. Murthy, et al. Data warehousing and analytics infrastructure atFacebook[C]. In Proceedings of the18th ACM SIGMOD Symposium2010, Indianaplis,2010,432-443
[54] R. Buyya, T. Corte, H. Jin. Petal: Distributed virtual disks[C]. In Proceeding of the HighPerformance Mass Storage and Parallel I/O: Technologies and Applications, Piscataway,2002,420-430
[55] F. Bek, M. F. Kaashoek, D. Karger, et al. Wide-Area cooperative storage with CFS[C]. InProceeding of the18th ACM Symp On Operating System Principles, Banff,2001,202-215
[56] J. A. Andy. A generalized replica placement strategy to optimize latency in a wide areadistributed storage system[C]. Proceedings of the2008International Workshop on Data-awareDistributed Computing, New York,2008,49-54
[57] Y. Ding, Y. Lu. Automatic data placement and replication in grids[C]. In Proc of the HIPC.Kochi,2009,30-39
[58] A. Adya, W. Bolosky, M. Castro, et al. Farsite: Federated, available, and reliable storage for anincompletely trusted environment[C]. Proceedings of the7th Symposium on OperatingSystems Design and Implementation, San Diego,2002,435-446
[59] R.Honicky, E. Miller. Replication under scalable hashing: A family of algorithms for scalabledecentralized data distribution[C]. Proceedings of the18th International Parallel&Distributed Processing, London,2004,216-231
[60] S. Weil, S. Brandt, E. Miller, et al. Grid resource management-crush: controlled, scalable,decentralized placement of replicated data [C]. Proceedings of the18th ACM Symposium OnOperating System Principles, USA,2006,221
[61] K. Shi, A. S.Thanamani. A new replica creation and placement algorithm for data gridenvironment[C]. Proceedings of the2010on Data Storage and Data Engineering, Bangalore,2010,265-269
[62] Q. Chen, B. Y. Zhang. Dynamic replica placement and location strategies for data grid[C].Proceedings of the2008on Computer Science and Software Engineering, Wuhan,2008,35-40
[63] A. S. Weil, A. B. Scott, L. Miller. Ceph: A scalable, high-performance distributed filesystem[C]. Proceedings of the7th Symposium on Operating Systems Design andImplementation, Santa Cruz,2006,307-320
[64] J. Baker, C. Bond, J. Corbett, et al. Megastore: providing scalable, highly available storage forinteractive services. Conference on Innovative Data Systems Research, Asilomar,2011,106-117
[65] M. Li, J. Su, W. Zheng. Grid Codes: Strip-based erasure codes with high fault tolerance forstorage systems [J]. ACM Transactions on Storage,2009,4(4):15:1-15:22
[66] G. Harihara，B. Janakiram，M. G. Chandra，et al. SpreadStore: A LDPC erasure code schemefor distributed storage system[C]. Proceedings of the2010International Conference on DataStorage and Data Engineering, Bangalore,2010,87-93
[67] M. K. Greenan，L. X. Zhou，J. J. Wylie. Flat XOR-based erasure codes in storage systemsConstructions, efficient recovery, and tradeoffs[C]. IEEE26th Symposium on Mass StorageSystems and Technologies, Incline Village,2010,1-14
[68] S. Reed, G. Solomon. Polynomial codes over certain finite fields [J]. Journal of the Society forIndustrial and Applied Mathematics,1960,8(2):300-304
[69] M. Rabin. Efficient dispersal of information for security loading balancing and fault tolerance[J]. Journal of the Association for Computing Machinery,1989,36(2):335-348
[70] Y. Chen, J. Edler, A. Goldberg, et al. A prototype implementation of archival intermemory[C].Fourth ACM Conference on Digital libraries Proceedings, New York,1999,28-37
[71] J. Plank. A tutorial on Reed-Solomon coding for fault-tolerance in RAID-like systems [J].Software-Practice and Experience,1997,27(9):995-1012
[72] J. Plank, Y. Ding. Note: Correction to the1997tutorial on Reed-Solomon coding [J].Software-Practice and Experience,2005,35(2):189-194
[73] R. M. Roth, A. Lempel. On MDS codes via Cauchy Matrices [J]. IEEE Transaction onInformation Theory,1989,35(6),1314-1319
[74] J. Blomer, M. Kalfane, M. Karpinski, et al. An XOR-based erasure-resilient coding scheme[R].Berkeley:International Computer Science Institute,1995,11-23
[75] J. Plank, L. Xu. Optimizing Cauchy Reed-Solomon codes for fault-tolerant network storageapplications[C]. IEEE International Symposium on Network Computing and ApplicationsProceedings, Cambridge,2006,173-180
[76] J. Plank, C. Schuman, B. Robison. Heuristics for optimizing matrix-based erasure codes forfault-tolerant storage systems[C]. International Conference on Dependable Systems andNetworks Proceedings, Boston,2012,1-12
[77] A. Dimakis, P. Godfrey, J. Wainwright, et al. Network coding for distributed storage systems
[C].26th IEEE International Conference on Computer Communications Proceedings, Alaska,2007,200-208
[78] A. Dimakis, P. Godfrey, Y. Wu, et al. Network coding for distributed storage systems [J]. IEEETransactions on Information Theory,2010,56(9):4539-4551
[79] A. Dimakis, K. Ramchandran, Y. Wu, et al. A survey on network codes for distributed storage
[C]. Proceedings of the IEEE, Carolina,2011,476-489
[80]常乾,许胤龙,项利萍,等.基于EVENODD码的单盘故障快速恢复算法[J].计算机应用与软件,2011,28(6):15-18
[81] L. Xiang, Y. Xu, C. S. Lui, et al. Optimal recovery of single disk failure in RDP code storagesystems [J]. ACM SIGMETRICS Performance Evaluation Review,2010,38(1):119-130
[82] K. V. Rashmi, N. B. Shah, P. V. Kumar, et al. Explicit construction of optimal exactregenerating codes for distributed storage [J]. Communication, Control,and Computing,2009,22(6):211-243
[83] Y. Hu, Y. Xu, X. Wang, et al. Cooperative recovery of distributed storage system frommultiple losses with network coding[J]. IEEE Journal on Selected Areas in Communication,2010,28(2):268-276
[84] L. You, K. Pollack, D. Long, Deep store: An archival storage system architecture[C].Proceedings of the21st IEEE International Conference on Data Engineering, California,2005,804-815
[85] S.Quinlan，S.Dorward, Venti: A new approach to archival storage[C]. In Proceedings of the2002Conference on File and Storage Technologies (FAST), California,2002,89-101
[86] M. Lillibridge, K. Eshghi, D. Bhagwat, et al. Sparse indexing: Large scale, inlinededuplication using sampling and locality [R]. Proceeding of the7thUSENIX Conference onFile and Storage Technologies, San Francisco,2009,111-123
[87] A. Mut, C. Benjie, D. Mazières, et al. A low bandwidth network file system [C] Proceeding ofthe18thACM Symposium on Principles, New York,2001,174-187
[88] M. O. Rabin. Fingerprinting by random polynomials[R]. Boston: Centerfor Research inComputing Technology, Harvard University,1981,23-34
[89] C. Y. Liu, Y. P. Lu, D David, etal. ADMAD: Application-driven metadata aware de-duplicationarchival storage system [C]. Proceeding of the25thIEEE Conf Mass Storage Syst and Technol.Piscataway,2008,29-35
[90]敖莉，舒继武，李明强.重复数据删除技术[J]，计算机研究与发展，2010,21(5):916-929
[91] W. E. Binary. HP storage works D2D4000backup system: a report and full Performance teston Hewlett-Packard's SME data deduplication appliance[R]. California: Hewlett-PackardDevelopment Company, April,2008
[92] L.Aronovich, R.Asher, E.Bachmat, et al. The design of a similarity based deduplicationsystem[C]. In Proceedings of SYSTOR2009:The Israeli Experimental Systems Conference，Israeli,2009,11-26
[93] EMC Corporation. EMC centera: content addressed storage system[R]. Massachusetts: EMCCorporation, May,2002
[94] H. Li, H. Patterson. Avoiding the disk bottleneck in the data domain deduplication filesystem[C]. Proceedings of the6th USENIX Conference on File And Storage Technologies,California,2008,1-14
[95] A. Adya, W. J. Bolosky, M. Castro, et al. FARSITE: Federated, available, and reliable storagefor an incompletely trusted environment[C]. Proceedings of the5th Symposium On OperatingSystems Design and Implementation (OSDI2002), New York,USA: ACM Press,2002,1-14
[96] W. J. Losky, S. Corbin, D. Goebel, et al. Single instance storage in Windows2000[C].Proceedings of the4th Usenix Windows System, Berkeley,2000,13-24
[97] P. Sehgal, V. Tarasov, E. Zadok. Optimizing energy and performance for server-class filesystem workloads [J]. IEEE Trans on Storage,2010,10(6):10-31
[98] J. Kim, D. Rotem. Energy proportionality for disk storage using replication[C]. In Proceedingsof the14th International Conference on Extending Database Technology, New York,2011:81-92
[99] E. Otoo, D. Rotem, S. C. Tsao. Analysis of trade-off between power saving and response timein disk storage systems[C]. In Proceedings of IEEE International Symposium on Parallel andDistributed Processing, Berkeley,2009,1012-1031
[100] H. Huang, W. D. Huang, G. K. Shin. FS2: Dynamic data replication in free disk space forimproving disk performance and energy Consumption [C] Proceedings of the20th ACMSymposium on Operating Systems Principles, Brighton,2005:263-276
[101] S. Gurumurthi, A. Sivasub, K. M. Demir, et al. DRPM:Dynamic speed control for powermanagement in server class disks [C] The ACM Proceedings of the International Symposiumon Computer Architecture, New York,2003,169-181
[102] S. Sankar, S. Gurumurthi, M. R.Stan. Intra-disk parallelism: An idea whose time has come [R].Virginia: University of Virginia, February,2008
[103] T. Bisson, S. Brandt, D. Long. A hybrid disk-aware spin-down algorithm with I/O subsystemsupport[C]. In Proceedings of the26th IEEE International Performance Computing andCommunications Conference, Los Alamitos,2007,236-245
[104] H. J. Lee, K. H. Lee, S. H. Noh. Augmenting raid with an ssd for energy relief[C]. InProceedings of the2008Conference on Power Aware Computing and Systems, Berkeley,2008,12-18
[105] D. Narayanan, E. Thereska, M. Donnelly, et al. Migrating server storage to ssds: analysis oftrades[C]. Proceedings of the4th ACM European Conference on Computer systems, NewYork,2009,145-158
[106] Y. Deng. What is the future of disk drives, death or rebirth?[J]. ACM Computing Surveys,2011,42(23):123-27
[107] E. Thereska, A. Donnelly, D. Narayanan. Sierra: practical power-proportionality for datacenter storage[C]. In Proceedings of the Sixth Conference on Computer Systems, New York,2006,169-182
[108]万哲先.代数和编码[M].北京:高等教育出版社,2007,201-234
[109]冯克勤.纠错码的代数理论[M].北京:清华大学出版社,2005,14-22
[110] S. Roman. Coding and information theory [M]. New York: Springer-verlag,1992,254-263
[111] J. S. Plank. XORs Lower bounds and MDS codes for storage[R]. Tennessee: University ofTennessee, May,2010
[112]董欢庆,李战怀,林伟. RAID-VCR:一种能够承受三个磁盘故障的RAID结构[J].计算机学报,2006,29(3):792-800
[113]魏青松.大规模分布式存储技术研究[D].成都：电子科技大学，2004
[114]周旭.面向Internet的大规模分布式存储技术研究[D].成都：电子科技大学，2004
[115]吴素贞.磁盘阵列高可用技术研究[D].武汉：华中科技大学，2010
[116]李洁琼.海量存储资源管理关键技术研究[D].武汉：华中科技大学，2010
[117]刘刚.分布式存储系统中基于纠错码的容错机制研究[D].武汉：华中科技大学，2007
[118] M. Li, J. Shu. Preventing silent data corruptions from propagating during data reconstruction[J]. IEEE Transactions on Computers,2010,59(12):1611-1624.
[119] M. Li, J. Shu. DACO: High-performance disk architecture designed especially for large scaleerasure-coded storage systems [J]. IEEE Transactions on Computers,2010,59(10):1350-1362.
[120] M. Li, J. Shu. On the equivalence between the B-Code constructions and perfectone-factorizations[C]. Proceedings of the2010IEEE International Symposium on InformationTheory (ISIT’10), Austin,2010,993-996.
[121] M. Li, J. Shu. On cyclic lowest density MDS array codes constructed using starters.Proceedings of the2010IEEE International Symposium on Information Theory (ISIT’10),Austin,2010,1315-1319.
[122] P. Corbett, B. English, A.Goel, et al. Row diagonal parity for double disk failure correction[C].3rdUsenix Conference on File and Storage Technologies, San Francisco,2004,121-132