认知可重构的高效能Web服务体系结构研究
|
摘要
通常Web服务的能力依赖于通用计算平台性能的提高，但是Web服务本身的业务特征和流量规律却与通用计算机体系中以科学计算为中心的设计初衷不尽相同，因此，Web服务的性能和效能无法直接从通用计算机性能的提高中获得预期的进步。
在大规模数据中心（Internet Data Center，简称IDC）应用中，随着规模和负荷动态范围的增大，Web服务的高能耗问题更加严重，已经成为高效能Web服务发展的瓶颈。为了解决这个问题，人们采取了多核处理、缓存预取、前端加速等多种研究思路，也取得了一定的理论和实践成果。这些研究成果在一定条件下获得了改进，但却未从体系结构上解决如何使服务体系与承载系统相匹配的问题。本文从体系结构着手，提出了适合Web业务特征和流量特点的架构。论文进行了如下创新性的工作：
提出了基于认知的主动重构Web服务体系结构（Proactive Reconfigurable Web serviceArchitecture，简称PRWA）；提出了PRWA模型的形式化定义；给出了PRWA模型的运行机理；导出了PRWA模型与现有体系结构模型之间的演化规律；理论证明了PRWA模型是一种面向Web服务的高效能模型。
提出了混合可重构Web处理阵列（Hybrid Reconfigurable Web service Array，简称HRWA）；给出了基于Web服务运算处理规律的算粒特征；提出了算粒提取的形式化描述方法；设计了设计了相关的原理验证系统；证明了HRWA结构是一种面向Web服务高效能计算结构。
提出了面向Web服务的智能混合存储结构（Smart Hybrid Memory Architecture，简称SHMA）；基于Web服务的数据结构和存储访问特征，设计了面向Web服务的智能混合存储体系；通过对词频统计和流媒体这两类典型Web服务的验证，证明了SHMA结构是一种面向Web服务的高效存储结构。
设计并实现了一种基于PRWA模型的高效能Web验证平台，经过实测，验证了该平台的Web服务能效比达到普通服务器的十倍以上。
The improvement of Web services mainly relies on the advance of architecture for generalcomputing. However, the characteristics of Web service do not match the architecture optimizedfor general computing. It results in a little improvements of Web service than expected.
The throughput of Web service blows up over the Internet. The data center enlarges its scaleand introduces multiple layers to make service available with commodity server. Therefore itresults in a severe consequence–energy problem. Multi-core processor, prefetched cache, andfront accelerator are proposed as a part of solution to Web service. They relieve the energyproblem within a small extent. It is rare that Web service is researched from a point of view ofarchitecture. A partial acceleration cannot solve the energy trouble. It is reported that theutilization of CPU is lower than10%in data center.
Our research is driven by application. Web service is I/O, storage, and computationintensive with the analysis of several typical applications. The architecture for generalcomputation cannot accommodate the diversity of Web service applications. It is the root of lowefficiency. The aim is to find an efficienty archteture for Web service. Our contrib utions can besummarized as follows:
1. Proactive Reconfigurable Web service Architecture (PRWA) is proposed as a solution. Aformal model of PRWA is established. Its architecture and working mechanism are clarified. Thehigh-performance characteristics of the PRWA architecture are demonstrated. PRWA is a goodsolution to application-oriented high-performance computing and services, which has a strongtheoretical impacts and practical references.
2. HRWA (Hybrid Reconfigurable Web service Array) is proposed. The computationfeatures and computing particles of Web service are summarized. A framework is put forward toextract computing particles of Web service. Then a method called BBC is designed within theframework. Futhermore, a HRWA with a set of some typ ical Web particles is implemented onFPGA to verify our idea. It is proved to be efficient for Web service.
3. SHSA (smart hybrid storage architecture) is proposed based on the analysis of thecharacteristics of storage access. It is evaluated with two app lications, word count and streamingmedia service. The results show that SHSA is efficient for Web service.
4. A testbed is designed for evaluation based on PRWA model. It obtains a speedup ofperformance more than ten times than that of a common Web server with our experiments.
在大规模数据中心（Internet Data Center，简称IDC）应用中，随着规模和负荷动态范围的增大，Web服务的高能耗问题更加严重，已经成为高效能Web服务发展的瓶颈。为了解决这个问题，人们采取了多核处理、缓存预取、前端加速等多种研究思路，也取得了一定的理论和实践成果。这些研究成果在一定条件下获得了改进，但却未从体系结构上解决如何使服务体系与承载系统相匹配的问题。本文从体系结构着手，提出了适合Web业务特征和流量特点的架构。论文进行了如下创新性的工作：
提出了基于认知的主动重构Web服务体系结构（Proactive Reconfigurable Web serviceArchitecture，简称PRWA）；提出了PRWA模型的形式化定义；给出了PRWA模型的运行机理；导出了PRWA模型与现有体系结构模型之间的演化规律；理论证明了PRWA模型是一种面向Web服务的高效能模型。
提出了混合可重构Web处理阵列（Hybrid Reconfigurable Web service Array，简称HRWA）；给出了基于Web服务运算处理规律的算粒特征；提出了算粒提取的形式化描述方法；设计了设计了相关的原理验证系统；证明了HRWA结构是一种面向Web服务高效能计算结构。
提出了面向Web服务的智能混合存储结构（Smart Hybrid Memory Architecture，简称SHMA）；基于Web服务的数据结构和存储访问特征，设计了面向Web服务的智能混合存储体系；通过对词频统计和流媒体这两类典型Web服务的验证，证明了SHMA结构是一种面向Web服务的高效存储结构。
设计并实现了一种基于PRWA模型的高效能Web验证平台，经过实测，验证了该平台的Web服务能效比达到普通服务器的十倍以上。
The improvement of Web services mainly relies on the advance of architecture for generalcomputing. However, the characteristics of Web service do not match the architecture optimizedfor general computing. It results in a little improvements of Web service than expected.
The throughput of Web service blows up over the Internet. The data center enlarges its scaleand introduces multiple layers to make service available with commodity server. Therefore itresults in a severe consequence–energy problem. Multi-core processor, prefetched cache, andfront accelerator are proposed as a part of solution to Web service. They relieve the energyproblem within a small extent. It is rare that Web service is researched from a point of view ofarchitecture. A partial acceleration cannot solve the energy trouble. It is reported that theutilization of CPU is lower than10%in data center.
Our research is driven by application. Web service is I/O, storage, and computationintensive with the analysis of several typical applications. The architecture for generalcomputation cannot accommodate the diversity of Web service applications. It is the root of lowefficiency. The aim is to find an efficienty archteture for Web service. Our contrib utions can besummarized as follows:
1. Proactive Reconfigurable Web service Architecture (PRWA) is proposed as a solution. Aformal model of PRWA is established. Its architecture and working mechanism are clarified. Thehigh-performance characteristics of the PRWA architecture are demonstrated. PRWA is a goodsolution to application-oriented high-performance computing and services, which has a strongtheoretical impacts and practical references.
2. HRWA (Hybrid Reconfigurable Web service Array) is proposed. The computationfeatures and computing particles of Web service are summarized. A framework is put forward toextract computing particles of Web service. Then a method called BBC is designed within theframework. Futhermore, a HRWA with a set of some typ ical Web particles is implemented onFPGA to verify our idea. It is proved to be efficient for Web service.
3. SHSA (smart hybrid storage architecture) is proposed based on the analysis of thecharacteristics of storage access. It is evaluated with two app lications, word count and streamingmedia service. The results show that SHSA is efficient for Web service.
4. A testbed is designed for evaluation based on PRWA model. It obtains a speedup ofperformance more than ten times than that of a common Web server with our experiments.
引文
[1] Verisign. Over7Million Domain Names Registered in Q22012, Total Domains Pass240MillionWorldwide [EB/OL].: CircleID Internet Infrastructure/Industry,2012-12-17:http://www.circleid.com/posts/20121002_over_7_million_domain_names_registered_in_q2_2012.
[2]王萌.全球服务器数量排行榜：谁家Web服务器最多[EB/OL].: IT经理网动态，数据中心,2012-03-12.:http://www.ctocio.com/ccnews/4820.html.
[3] Dung H., Phan, Junichi Suzuki, and Raymond Carrollet al. Evolutionary multiobjective optimization forgreen clouds [M]. ACM，2012:19-26.
[4] Yang Cao, Jienan Cao, Zeyu Xiong. Research and Implementation on Resource Saving-Based ServerCluster Scheduling System [J]. Advances in Automation and Robotics, Vol.2.2012:115-120
[5]李旭阳.数据中心算笔节能账[EB/OL].:ZCOM计算机世界，2012.49，2012：http://www.zcom.com/article/93767/
[6]孙永杰.惠普助力绿色数据中心专注情境感知[J].通信世界，2012，（44）:21.
[7]陈康，郑纬民.云计算:系统实例与研究现状[J].软件学报,2009,20（5）:1337-1348.
[8]李琼，汪审权，方粮等. SAN存储技术研究[J].计算机工程,2003,19:165-167.
[9] Paul Barham, Boris Dragovic, Keir Fraseret al. Xen and the art of virtualization [J]. ACM SIGOPSOperating Systems Review,2003,37（5）:164-177.
[10] Michael Armbrust, Armando Fox, Rean Griffithet al. A view of cloud computing [J]. Communications ofthe ACM,2010,53（4）:50-58.8
[11]孙利民，沈杰.从云计算到海计算:论物联网的体系结构[J]. ZTE TECHNOLOGY JOURNAL,2011:9.
[12]张建勋，古志民，郑超.云计算研究进展综述[J].计算机应用研究,2010,27（2）:429-433.
[13]工信部.工业节能―十二五‖规划[M].北京：工信部，2012:27.
[14]李国杰.信息科学技术的长期发展趋势和我国的战略取向[J].中国教育网络,2010,（002）:27-30.
[15] Wei-Tek Tsai, Xin Sun, Janaka Balasooriya. Service-oriented cloud computing architecture [M]. IEEE，2010:684-689.
[16] Ian Foster, Yong Zhao, Ioan Raicuet al. Cloud computing and grid computing360-degree compared [M].Ieee，2008:1-10.
[17]Qi Zhang, Lu Cheng, Raouf Boutaba. Cloud computing: state-of-the-art and research challenges [J].Journal of Internet Services and Applications.2010,1（1）:7-18.
[18] Liang-Jie Zhang,Qun Zhou. CCOA: Cloud computing open architecture [M]. Ieee，2009:607-616.
[19] Rajkumar Buyya, Rajiv Ranjan, Rodrigo Calheiros. Intercloud: Utility-oriented federation of cloudcomputing environments for scaling of application services [J]. Algorithms and architectures for parallelprocessing,2010:13-31.
[20] Quan Z. Sheng, Boualem Benatallah, Zakaria Maamaret al. Configurable composition and adaptiveprovisioning of web services [J]. Services Computing, IEEE Transactions on,2009,2（1）:34-49.
[21] Amazon Elastic Compute Cloud. Amazon web services [J]. Retrieved November,2011,9:2011.
[22]杜晓梅，王广益，李利.最新TOP500超级计算机分析[J]. Red,2010,458（163840）:10.
[23] Rajkumar Buyya. High Performance Cluster Computing: Architectures and Systems (Volume1)[J].Prentice Hall, Upper SaddleRiver, NJ, USA,1999,1:999.
[24] Rob Armstrong, Dennis Gannon, Al Geistet al. Toward a common component architecture forhigh-performance scientific computing [M] IEEE，1999:115-124.
[25] Horacio Andrés Lagar-Cavilla, Joseph Andrew Whitney, Adin Matthew Scannellet al. SnowFlock: rapidvirtual machine cloning for cloud computing [M] ACM,2009:1-12.
[26] Jason S. Reich, Matthew Naylor, Colin Runciman. Super compilation and the Reduceron [M].2010.
[27] Philip HW Leong. Recent trends in FPGA architectures and applications [M]. IEEE，2008:137-141.
[28] Xiaochen Guo, Engin Ipek, Tolga Soyata. Resistive computation: avoiding the power wall withlow-leakage, STT-MRAM based computing [M]. ACM,2010:371-382.
[29]张莉.基于FPGA的行波管夹持杆自动测试系统时序控制卡研制[D].成都:电子科技大学，2011
[30]齐冶.基于微处理器的FPGA配置方案[EB/OL].:电子创新网方案，2010-3-18：http://www.eetrend.com/solution/100024598.
[31]曾繁泰，侯亚宁，可编程器件等.可编程器件应用导论[M]:清华大学出版社，2001.
[32]安杰，胡飞.形式计算的计算量与信息熵的关系以及在算法优化中的应用[J].航空计算技术,2007,37（1）:58-60.
[33]赵楠.一种基于网格与R树的多级混合索引[J].计算机技术,2009,19（3）:91-94.
[34]文小明. Web2.0环境下的信息服务[J].湘潭师范学院学报:社会科学版,2006,28（6）:212-214.
[35]王鹏.通用多核集群上的并行调优策略研究[D].上海:复旦大学，2011.
[36] Tim O Reilly. Web2.0: compact definition [EB/OL].: O'reilly/News&Commentary,2005:http://radar.oreilly.com/archives/2005/10/web_20_compact_definition.html.
[37] Erik Christensen, Francisco Curbera, Greg Meredithet al. Web services description language (WSDL)1.1
[M],2001.
[38] Tim Bray, Jean Paoli, C. Michael Sperberg-McQueenet al. Extensible markup language (XML)[J]. WorldWide Web Journal,1997,2（4）:27-66.
[39] Francisco Curbera, Matthew Duftler, Rania Khalafet al. Unraveling the Web services web: anintroduction to SOAP, WSDL, and UDDI [J]. Internet Computing, IEEE,2002,6（2）:86-93.
[40] Karl Czajkowski, Steven Fitzgerald, Ian Fosteret al. Grid information services for distributed resourcesharing [M] IEEE，2001:181-194.
[41] Mitsuhisa Sato,Hidemoto Nakada,Satoshi Sekiguchiet al. Ninf: A network based information library forglobal world-wide computing infrastructure[M]Springer，1997:491-502.
[42] Ronald P. Doyle, Jeffrey S. Chase, Omer Asadet al. Model-based resource provisioning in a web serviceutility [M]，2003:5.
[43] Steve Graham, Anish Karmarkar, Jeff Mischkinskyet al. Web services resource1.2(ws-resource)[J].OASIS Standard，April.2006.
[44] Mike P. Papazoglou. Service-oriented computing: Concepts, characteristics and directions [M]. IEEE，2003:3-12.
[45] Daniel A. Menasce, Virgilio AF Almeida. Capacity Planning for Web Services: metrics, models, andmethods[M]: Prentice Hall Upper Saddle River，2002.
[46] Lada A. Adamic, Bernardo A. Huberman. Zipf’s law and the Internet[J]. Glottometrics,2002,3（1）:143-150.
[47] Atif Shahab,Danny Chuon,Toyotaro Suzumuaet al. Grid portal interface for interactive use andmonitoring of high-throughput proteome annotation[J]. Grid Computing in Life Science,2005:53-67.
[48]Mudashiru Busari,Carey Williamson. ProWGen: a synthetic workload generation tool for simulationevaluation of web proxy caches[J]. Computer Networks.2002.38（6）:779-794
[49] He Lingjuan, Liu Lianchen, Wu Cheng. Asymmetric Web service operation matching in structural-levelsimilarity measure[M]IEEE，2008:688-693.
[50]陶永才. Web访问特征模拟器的设计与实现[D].郑州:郑州大学，2005
[51]王会霞，叶海琴. Web访问特征模型建模[J].电脑知识与技术,2010,6（8）:1804-1806.
[52]许欢庆，王永成.基于用户访问路径分析的网页预取模型[J].软件学报,2003.14（6）:1142-1147.
[53] B. A. Nayfeh, K. Olukotun. A single-chip multiprocessor [J]. Computer,1997,30（9）:79-85.
[54] Nathan Tuck, Dean M. Tullsen. Initial observations of the simultaneous multithreading Pentium4processor [M]. IEEE，2003:26-34.
[55] LTD Tilera. Tile processor architecturetechnology brief [EB/OL].: Tilera,2007: http://www.tilera.com.
[56] Ashish Dixit. Networking applications for xtensa configurable processors [J]. Linley Tech,2006.
[57] Wm A. Wulf, Sally A. McKee. Hitting the memory wall: Implications of the obvious [J]. ACMSIGARCH Computer Architecture News,1995,23（1）:20-24.
[58] Hiroshi Inoue, Hideaki Komatsu, Toshio Nakatani. A study of memory management for web-basedapplications on multicore processors [J]. ACM Sigplan Notices,2009,44（6）:386-396.
[59] Gary Lauterbach. Recovery Act: SeaMicro Volume Server Power Reduction Research Development[R]:SeaMicro，2012
[60] Anil Rao. SeaMicro technology overview[R]: Tech. rep., SeaMicro，2010
[61] WinCom. Enhancing Web Performance with the WinCom Switching Server and Storage Area Networks
[EB/OL].:百度文库：2008. http://wenku.baidu.com/view/7fde3e88680203d8ce2f243f.html?from=related
[62] Christine Burns, David Newman. Vendors choose to tout disparate application-acceleration techniques [J].Network World,2006,1:16.
[63] Hung-Cheng Huang, Ming-Che Yu, Tomonori Hirai. Flexible network switch fabric for clustering system
[M]. Google Patents，2012.
[64] Wei Lu, Jamshid Mahdavi, Darrell Long. CONTENT ASSOCIATIVE CACHING METHOD FOR WebAPPLICATIONS [M]. Google Patents，2010
[65] P. Hunter. Chipping away at long-distance data[J]. Engineering&Technology,2012,7（8）:78-81.
[66] Tralvex Yeap. USER MOVEMENT INTERPRETATION IN COMPUTER GENERATED REALITY
[M]. Google Patents，2010
[67] Robert Graybill, Rami Melhem. Power aware computing[M]: Kluwer Academic/Plenum，2002
[68] James B. Lee, Brent Ware. Open Source Web Development with LAMP: Using Linux, Apache, MySQL,Perl, and PHP[M]: Addison-Wesley Professional，2003
[69]马骁，徐钊.基于多核网络处理器平台的Libpcap性能评估[J]:
[70] Liang Liu, Hao Wang, Xue Liuet al. GreenCloud: a new architecture for green data center [M]. ACM，2009:29-38.
[71] Klaus-Dieter Lange. Identifying shades of green: The SPECpower benchmarks [J]. Computer,2009,42（3）:95-97.
[72]吴军.浪潮之巅[M]:电子工业出版社，2011
[73]韩燕波，王桂玲，刘晨.网格、云、SaaS、SOA、WebX.0——本质是互联网计算[J].中国计算机学会通讯,2009,11（5）:55-58.
[74] Yale N. Pott, Robert G. Sheldon, Michael Shebanowet al. A Comparison of Several Evolving (University)Supercomputer Architectures[M]IEEE Computer Society，1984:15.
[75] Kenneth R. Wallgren. Optical Disks and Supercomputers [M]. International Society for Optics andPhotonics，1985:212-216.
[76]徐新海，杨学军，林宇斐等.一种面向CPU-GPU异构系统的容错方法[J].软件学报,2011,22（10）:2538-2552.
[77]朱平，朱建涛，高剑刚等.高性能计算存储关键技术研究[J]. Journal of Computer Research andDevelopment,2011,48:354-364.
[78] Thomas Worsch. Programming environments for cellular automata [M]. Springer，1997
[79] Leslie G. Valiant. A bridging model for parallel computation[J]. Communications of the ACM,1990,33（8）:103-111.
[80] Steven Fortune, James Wyllie. Parallelism in random access machines [M]. ACM，1978:114-118.
[81] Stephen Cook, Cynthia Dwork, Rüdiger Reischuk. Upper and lower time bounds for parallel randomaccess machines without simultaneous writes [J]. SIAM Journal on Computing,1986,15（1）:87-97.
[82] Richard Cole, Ofer Zajicek. The APRAM: Incorporating asynchrony into the PRAM model [M]. ACM，1989:169-178.
[83] Albert Alexandrov, Mihai F. Ionescu, Klaus E. Schauseret al. LogGP: Incorporating long messages intothe LogP model for parallel computation [J]. Journal of parallel and distributed computing,1997,44（1）:71-79.
[84] Pat Morin. Coarse grained parallel computing on heterogeneous systems [M]. ACM，1998:628-634.
[85] Tiffani L. Williams, Rebecca J. Parsons. The heterogeneous bulk synchronous parallelmodel[M]Springer，2000:102-108.
[86] JI Yong Chang DING Qun,CHEN Guo Liang AN Hong. A Realistic Parallel Computational Model [J].Chinese Journal of Computers,2001,4:15
[87] Michael Butts, Anthony Mark Jones, Paul Wasson. A structural object programming model, architecture,chip and tools for reconfigurable computing [M]. IEEE，2007:55-64.
[88] Eric E. Johnson. Completing an MIMD multiprocessor taxonomy [J]. ACM SIGARCH ComputerArchitecture News,1988,16（3）:44-47.
[89] Kai Hwang, Tian-pong Chang. Combinatorial reliability analys is of multiprocessor computers [J].Reliability, IEEE Transactions on,1982,31（5）:469-473.
[90] Howard Jay Siegel. Interconnection networks for SIMD machines[J]. Computer,1979,12（6）:57-65.
[91] Arne Halaas, B rge Svingen, Magnar Nedlandet al. A recursive MISD architecture for pattern matching[J]. Very Large Scale Integration (VLSI) Systems, IEEE Transactions on,2004,12（7）:727-734.
[92] Burton J. Smith. A pipelined, shared resource MIMD computer[M].IEEE Computer Society Press，1986:39-41.
[93] M. Woodside, John E. Neilson, Dorina C. Petriuet al. The stochastic rendezvous network model forperformance of synchronous client-server-like distributed software [J]. Computers, IEEE Transactions on,1995,44（1）:20-34.
[94] Yu Dan, Chen XinMeng, Chang YunLei. An improved P2P model based on Chord[M].IEEE，2005:807-811.
[95] Juraj Wiedermann. Parallel Turing machines [J]. Techn. Rep. RUU-CS-84-11, Dept. of Computer Science,University of Utrecht, Utrecht.1984:1-35.
[96]邬江兴，罗兴国.基于认知的主动重构计算体系[M].上海:中国工程院高端论坛论文集，2012
[97] Arthur S. Bland, Ricky A. Kendall, Douglas B. Kotheet al. Jaguar: The world’s most powerful computer[J]. Memory (TB),2009,300（62）:362.
[98] Kevin J. Barker, Kei Davis, Adolfy Hoisieet al. Entering the petaflop era: the architecture andperformance of Roadrunner[M].IEEE Press，2008:1.
[99]蒋本珊.计算机组成原理[M]:清华大学出版社，2004
[100]王一宾，韩承双.计算机体系结构与软件体系结构[J].安徽建筑工业学院学报(自然科学版),2005,1:20.
[101]黄海，吴昊.基于CUDA的立体匹配算法评估工具研究[J].系统仿真学报,2012,24（001）:154-157.
[102] Cristiana Amza, Alan L. Cox, Sandhya Dwarkadaset al. Treadmarks: Shared memory computing onnetworks of workstations [J]. Computer,1996,29（2）:18-28.
[103] Reiner Hartenstein. A decade of reconfigurable computing: a visionary retrospective [M]. IEEE Press，2001:642-649.
[104] David Kielpinski, Chris Monroe,David J. Wineland. Architecture for a large-scale ion-trap quantumcomputer [J]. Nature Publishing Group,2002:709-711.
[105] Forest Baskett, Alan Jay Smith. Interference in multiprocessor computer systems with interleavedmemory [J]. Communications of the ACM,1976,19（6）:327-334.
[106] Tim J. Todman, George A. Constantinides, Steven JE Wiltonet al. Reconfigurable computing:architectures and design methods [M].IET，2005:193-207.
[107] Malcolm Ware, Karthick Rajamani, Michael Floydet al. Architecting for power management: TheIBM POWER7approach [M]. IEEE，2010:1-11.
[108] B. Sinharoy, R. Kalla, W. J. Starkeet al. IBM POWER7multicore server processor [J]. IBM Journal ofResearch and Development,2011,55（3）:1.
[109] Victor W. Lee, Changkyu Kim, Jatin Chhuganiet al. Debunking the100X GPU vs. CPU myth: anevaluation of throughput computing on CPU and GPU[M].ACM，2010:451-460.
[110] Johannes Gilger, Johannes Barnickel, Ulrike Meyer. GPU-Acceleration of Block Ciphers in theOpenSSL Cryptographic Library [J]. Information Security,2012:338-353.
[111] InfiniBand Trade Association. InfiniBand Architecture Specification: Release1.0[M]. InfiniBand TradeAssociation，2000.
[112] Jiuxing Liu, Jiesheng Wu, Dhabaleswar K. Panda. High performance RDMA-based MPI implementationover InfiniBand [J]. International Journal of Parallel Programming,2004,32（3）:167-198.
[113] Eugene L. Lawler, Jan Karel Lenstra, AHG Rinnooy Kanet al. The traveling salesman problem: a guidedtour of combinatorial optimization[M].Wiley Chichester，1985.
[114]李弋.云环境中典型应用的I/O优化策略研究[D].上海:复旦大学，2012
[115] Peter M. Kogge. EXECUBE-a new architecture for scaleable MPPs[M]IEEE，1994:77-84
[116] David Patterson, Thomas Anderson,Neal Cardwellet al. A case for intelligent RAM[J]. Micro,IEEE,1997,17（2）:34-44.
[117] Ashish A. Pandya. Complex symbol evaluation for programmable intelligent search memory[M].GooglePatents，2010.
[118] Todd M. Austin. DIVA: A reliable substrate for deep submicron microarchitecture design [M]. IEEE，1999:196-207.
[119] Seung-Moon Yoo, Jose Renau, Michael Huanget al. Flexram architecture design parameters [J]. Centerfor Supercomputing Research and Development (CSRD), Tech, Rep.2000:1584.
[120] Mark Oskin, Frederic T. Chong, Timothy Sherwood. Active pages: A computation model for intelligentmemory [M]: IEEE Computer Society，1998
[121] Jack J. Dongarra, David W. Walker. The quest for petascale computing[J]. Computing in Science&Engineering,2001,3（3）:32-39.
[122] Steve Leibson, James Kim. Configurable processors: a new era in chip design [J]. Computer,2005,38（7）:51-59.
[123] Peter M. Chen, Edward K. Lee,Garth A. Gibsonet al. RAID: High-performance, reliable secondarystorage[J]. ACM Computing Surveys (CSUR),1994,26（2）:145-185.
[124] Jim Gray. Tape is dead, disk is tape, flash is disk, RAM locality is king[EB/OL].:Gong ShowPresentation at CIDR,2006-12-10
[125] Erik Riedel, Garth Gibson, Christos Faloutsos. Active storage for large-scale data mining andmultimedia applications[M].Citeseer，1998:62-73.
[126] Parthasarathy Ranganathan. From microprocessors to nanostores: Rethinking system building blocks forthe data-centric era [J],2011:1.
[127]Kevin Lim,Parthasarathy Ranganathan,Jichuan Changet al. Understanding and des igning new serverarchitectures for emerging warehouse-computing environments [M]. IEEE，2008:315-326
[128] Kevin Lim, Parthasarathy Ranganathan, Jichuan Changet al. Understanding and designing new serverarchitectures for emerging warehouse-computing environments [M]. IEEE，2008:315-326.
[129] David G. Andersen, Jason Franklin, Michael Kaminskyet al. FAWN: A fast array of wimpy nodes [M].ACM，2009:1-14.
[130] Behrouz A. Forouzan. TCP/IP protocol suite [M]. McGraw-Hill, Inc.，2002.
[131]Shen Gao Jianliang Xu, Bingsheng He, Byron Choi Haibo Hu. PCMLogging: Reducing TransactionLogging Overhead with PCM [EB/OL].2011:https://www.comp.hkbu.edu.hk/~xujl/Papers/cikm2011_gao.pdf
[132] Cristiano P. Costa, Italo S. Cunha, Alex Borgeset al. Analyzing client interactivity in streaming media
[M]. ACM，2004:534-543.
[133] Lei Guo, Songqing Chen, Zhen Xiaoet al. DISC: Dynamic interleaved segment caching for interactivestreaming [M]. IEEE，2005:763-772.
[134]王国权，李弋，叶德建.针对IPTV快进快退优化的媒体封装格式[J]. Computer Engineering,2010,36（6）:218-220.
[2]王萌.全球服务器数量排行榜：谁家Web服务器最多[EB/OL].: IT经理网动态，数据中心,2012-03-12.:http://www.ctocio.com/ccnews/4820.html.
[3] Dung H., Phan, Junichi Suzuki, and Raymond Carrollet al. Evolutionary multiobjective optimization forgreen clouds [M]. ACM，2012:19-26.
[4] Yang Cao, Jienan Cao, Zeyu Xiong. Research and Implementation on Resource Saving-Based ServerCluster Scheduling System [J]. Advances in Automation and Robotics, Vol.2.2012:115-120
[5]李旭阳.数据中心算笔节能账[EB/OL].:ZCOM计算机世界，2012.49，2012：http://www.zcom.com/article/93767/
[6]孙永杰.惠普助力绿色数据中心专注情境感知[J].通信世界，2012，（44）:21.
[7]陈康，郑纬民.云计算:系统实例与研究现状[J].软件学报,2009,20（5）:1337-1348.
[8]李琼，汪审权，方粮等. SAN存储技术研究[J].计算机工程,2003,19:165-167.
[9] Paul Barham, Boris Dragovic, Keir Fraseret al. Xen and the art of virtualization [J]. ACM SIGOPSOperating Systems Review,2003,37（5）:164-177.
[10] Michael Armbrust, Armando Fox, Rean Griffithet al. A view of cloud computing [J]. Communications ofthe ACM,2010,53（4）:50-58.8
[11]孙利民，沈杰.从云计算到海计算:论物联网的体系结构[J]. ZTE TECHNOLOGY JOURNAL,2011:9.
[12]张建勋，古志民，郑超.云计算研究进展综述[J].计算机应用研究,2010,27（2）:429-433.
[13]工信部.工业节能―十二五‖规划[M].北京：工信部，2012:27.
[14]李国杰.信息科学技术的长期发展趋势和我国的战略取向[J].中国教育网络,2010,（002）:27-30.
[15] Wei-Tek Tsai, Xin Sun, Janaka Balasooriya. Service-oriented cloud computing architecture [M]. IEEE，2010:684-689.
[16] Ian Foster, Yong Zhao, Ioan Raicuet al. Cloud computing and grid computing360-degree compared [M].Ieee，2008:1-10.
[17]Qi Zhang, Lu Cheng, Raouf Boutaba. Cloud computing: state-of-the-art and research challenges [J].Journal of Internet Services and Applications.2010,1（1）:7-18.
[18] Liang-Jie Zhang,Qun Zhou. CCOA: Cloud computing open architecture [M]. Ieee，2009:607-616.
[19] Rajkumar Buyya, Rajiv Ranjan, Rodrigo Calheiros. Intercloud: Utility-oriented federation of cloudcomputing environments for scaling of application services [J]. Algorithms and architectures for parallelprocessing,2010:13-31.
[20] Quan Z. Sheng, Boualem Benatallah, Zakaria Maamaret al. Configurable composition and adaptiveprovisioning of web services [J]. Services Computing, IEEE Transactions on,2009,2（1）:34-49.
[21] Amazon Elastic Compute Cloud. Amazon web services [J]. Retrieved November,2011,9:2011.
[22]杜晓梅，王广益，李利.最新TOP500超级计算机分析[J]. Red,2010,458（163840）:10.
[23] Rajkumar Buyya. High Performance Cluster Computing: Architectures and Systems (Volume1)[J].Prentice Hall, Upper SaddleRiver, NJ, USA,1999,1:999.
[24] Rob Armstrong, Dennis Gannon, Al Geistet al. Toward a common component architecture forhigh-performance scientific computing [M] IEEE，1999:115-124.
[25] Horacio Andrés Lagar-Cavilla, Joseph Andrew Whitney, Adin Matthew Scannellet al. SnowFlock: rapidvirtual machine cloning for cloud computing [M] ACM,2009:1-12.
[26] Jason S. Reich, Matthew Naylor, Colin Runciman. Super compilation and the Reduceron [M].2010.
[27] Philip HW Leong. Recent trends in FPGA architectures and applications [M]. IEEE，2008:137-141.
[28] Xiaochen Guo, Engin Ipek, Tolga Soyata. Resistive computation: avoiding the power wall withlow-leakage, STT-MRAM based computing [M]. ACM,2010:371-382.
[29]张莉.基于FPGA的行波管夹持杆自动测试系统时序控制卡研制[D].成都:电子科技大学，2011
[30]齐冶.基于微处理器的FPGA配置方案[EB/OL].:电子创新网方案，2010-3-18：http://www.eetrend.com/solution/100024598.
[31]曾繁泰，侯亚宁，可编程器件等.可编程器件应用导论[M]:清华大学出版社，2001.
[32]安杰，胡飞.形式计算的计算量与信息熵的关系以及在算法优化中的应用[J].航空计算技术,2007,37（1）:58-60.
[33]赵楠.一种基于网格与R树的多级混合索引[J].计算机技术,2009,19（3）:91-94.
[34]文小明. Web2.0环境下的信息服务[J].湘潭师范学院学报:社会科学版,2006,28（6）:212-214.
[35]王鹏.通用多核集群上的并行调优策略研究[D].上海:复旦大学，2011.
[36] Tim O Reilly. Web2.0: compact definition [EB/OL].: O'reilly/News&Commentary,2005:http://radar.oreilly.com/archives/2005/10/web_20_compact_definition.html.
[37] Erik Christensen, Francisco Curbera, Greg Meredithet al. Web services description language (WSDL)1.1
[M],2001.
[38] Tim Bray, Jean Paoli, C. Michael Sperberg-McQueenet al. Extensible markup language (XML)[J]. WorldWide Web Journal,1997,2（4）:27-66.
[39] Francisco Curbera, Matthew Duftler, Rania Khalafet al. Unraveling the Web services web: anintroduction to SOAP, WSDL, and UDDI [J]. Internet Computing, IEEE,2002,6（2）:86-93.
[40] Karl Czajkowski, Steven Fitzgerald, Ian Fosteret al. Grid information services for distributed resourcesharing [M] IEEE，2001:181-194.
[41] Mitsuhisa Sato,Hidemoto Nakada,Satoshi Sekiguchiet al. Ninf: A network based information library forglobal world-wide computing infrastructure[M]Springer，1997:491-502.
[42] Ronald P. Doyle, Jeffrey S. Chase, Omer Asadet al. Model-based resource provisioning in a web serviceutility [M]，2003:5.
[43] Steve Graham, Anish Karmarkar, Jeff Mischkinskyet al. Web services resource1.2(ws-resource)[J].OASIS Standard，April.2006.
[44] Mike P. Papazoglou. Service-oriented computing: Concepts, characteristics and directions [M]. IEEE，2003:3-12.
[45] Daniel A. Menasce, Virgilio AF Almeida. Capacity Planning for Web Services: metrics, models, andmethods[M]: Prentice Hall Upper Saddle River，2002.
[46] Lada A. Adamic, Bernardo A. Huberman. Zipf’s law and the Internet[J]. Glottometrics,2002,3（1）:143-150.
[47] Atif Shahab,Danny Chuon,Toyotaro Suzumuaet al. Grid portal interface for interactive use andmonitoring of high-throughput proteome annotation[J]. Grid Computing in Life Science,2005:53-67.
[48]Mudashiru Busari,Carey Williamson. ProWGen: a synthetic workload generation tool for simulationevaluation of web proxy caches[J]. Computer Networks.2002.38（6）:779-794
[49] He Lingjuan, Liu Lianchen, Wu Cheng. Asymmetric Web service operation matching in structural-levelsimilarity measure[M]IEEE，2008:688-693.
[50]陶永才. Web访问特征模拟器的设计与实现[D].郑州:郑州大学，2005
[51]王会霞，叶海琴. Web访问特征模型建模[J].电脑知识与技术,2010,6（8）:1804-1806.
[52]许欢庆，王永成.基于用户访问路径分析的网页预取模型[J].软件学报,2003.14（6）:1142-1147.
[53] B. A. Nayfeh, K. Olukotun. A single-chip multiprocessor [J]. Computer,1997,30（9）:79-85.
[54] Nathan Tuck, Dean M. Tullsen. Initial observations of the simultaneous multithreading Pentium4processor [M]. IEEE，2003:26-34.
[55] LTD Tilera. Tile processor architecturetechnology brief [EB/OL].: Tilera,2007: http://www.tilera.com.
[56] Ashish Dixit. Networking applications for xtensa configurable processors [J]. Linley Tech,2006.
[57] Wm A. Wulf, Sally A. McKee. Hitting the memory wall: Implications of the obvious [J]. ACMSIGARCH Computer Architecture News,1995,23（1）:20-24.
[58] Hiroshi Inoue, Hideaki Komatsu, Toshio Nakatani. A study of memory management for web-basedapplications on multicore processors [J]. ACM Sigplan Notices,2009,44（6）:386-396.
[59] Gary Lauterbach. Recovery Act: SeaMicro Volume Server Power Reduction Research Development[R]:SeaMicro，2012
[60] Anil Rao. SeaMicro technology overview[R]: Tech. rep., SeaMicro，2010
[61] WinCom. Enhancing Web Performance with the WinCom Switching Server and Storage Area Networks
[EB/OL].:百度文库：2008. http://wenku.baidu.com/view/7fde3e88680203d8ce2f243f.html?from=related
[62] Christine Burns, David Newman. Vendors choose to tout disparate application-acceleration techniques [J].Network World,2006,1:16.
[63] Hung-Cheng Huang, Ming-Che Yu, Tomonori Hirai. Flexible network switch fabric for clustering system
[M]. Google Patents，2012.
[64] Wei Lu, Jamshid Mahdavi, Darrell Long. CONTENT ASSOCIATIVE CACHING METHOD FOR WebAPPLICATIONS [M]. Google Patents，2010
[65] P. Hunter. Chipping away at long-distance data[J]. Engineering&Technology,2012,7（8）:78-81.
[66] Tralvex Yeap. USER MOVEMENT INTERPRETATION IN COMPUTER GENERATED REALITY
[M]. Google Patents，2010
[67] Robert Graybill, Rami Melhem. Power aware computing[M]: Kluwer Academic/Plenum，2002
[68] James B. Lee, Brent Ware. Open Source Web Development with LAMP: Using Linux, Apache, MySQL,Perl, and PHP[M]: Addison-Wesley Professional，2003
[69]马骁，徐钊.基于多核网络处理器平台的Libpcap性能评估[J]:
[70] Liang Liu, Hao Wang, Xue Liuet al. GreenCloud: a new architecture for green data center [M]. ACM，2009:29-38.
[71] Klaus-Dieter Lange. Identifying shades of green: The SPECpower benchmarks [J]. Computer,2009,42（3）:95-97.
[72]吴军.浪潮之巅[M]:电子工业出版社，2011
[73]韩燕波，王桂玲，刘晨.网格、云、SaaS、SOA、WebX.0——本质是互联网计算[J].中国计算机学会通讯,2009,11（5）:55-58.
[74] Yale N. Pott, Robert G. Sheldon, Michael Shebanowet al. A Comparison of Several Evolving (University)Supercomputer Architectures[M]IEEE Computer Society，1984:15.
[75] Kenneth R. Wallgren. Optical Disks and Supercomputers [M]. International Society for Optics andPhotonics，1985:212-216.
[76]徐新海，杨学军，林宇斐等.一种面向CPU-GPU异构系统的容错方法[J].软件学报,2011,22（10）:2538-2552.
[77]朱平，朱建涛，高剑刚等.高性能计算存储关键技术研究[J]. Journal of Computer Research andDevelopment,2011,48:354-364.
[78] Thomas Worsch. Programming environments for cellular automata [M]. Springer，1997
[79] Leslie G. Valiant. A bridging model for parallel computation[J]. Communications of the ACM,1990,33（8）:103-111.
[80] Steven Fortune, James Wyllie. Parallelism in random access machines [M]. ACM，1978:114-118.
[81] Stephen Cook, Cynthia Dwork, Rüdiger Reischuk. Upper and lower time bounds for parallel randomaccess machines without simultaneous writes [J]. SIAM Journal on Computing,1986,15（1）:87-97.
[82] Richard Cole, Ofer Zajicek. The APRAM: Incorporating asynchrony into the PRAM model [M]. ACM，1989:169-178.
[83] Albert Alexandrov, Mihai F. Ionescu, Klaus E. Schauseret al. LogGP: Incorporating long messages intothe LogP model for parallel computation [J]. Journal of parallel and distributed computing,1997,44（1）:71-79.
[84] Pat Morin. Coarse grained parallel computing on heterogeneous systems [M]. ACM，1998:628-634.
[85] Tiffani L. Williams, Rebecca J. Parsons. The heterogeneous bulk synchronous parallelmodel[M]Springer，2000:102-108.
[86] JI Yong Chang DING Qun,CHEN Guo Liang AN Hong. A Realistic Parallel Computational Model [J].Chinese Journal of Computers,2001,4:15
[87] Michael Butts, Anthony Mark Jones, Paul Wasson. A structural object programming model, architecture,chip and tools for reconfigurable computing [M]. IEEE，2007:55-64.
[88] Eric E. Johnson. Completing an MIMD multiprocessor taxonomy [J]. ACM SIGARCH ComputerArchitecture News,1988,16（3）:44-47.
[89] Kai Hwang, Tian-pong Chang. Combinatorial reliability analys is of multiprocessor computers [J].Reliability, IEEE Transactions on,1982,31（5）:469-473.
[90] Howard Jay Siegel. Interconnection networks for SIMD machines[J]. Computer,1979,12（6）:57-65.
[91] Arne Halaas, B rge Svingen, Magnar Nedlandet al. A recursive MISD architecture for pattern matching[J]. Very Large Scale Integration (VLSI) Systems, IEEE Transactions on,2004,12（7）:727-734.
[92] Burton J. Smith. A pipelined, shared resource MIMD computer[M].IEEE Computer Society Press，1986:39-41.
[93] M. Woodside, John E. Neilson, Dorina C. Petriuet al. The stochastic rendezvous network model forperformance of synchronous client-server-like distributed software [J]. Computers, IEEE Transactions on,1995,44（1）:20-34.
[94] Yu Dan, Chen XinMeng, Chang YunLei. An improved P2P model based on Chord[M].IEEE，2005:807-811.
[95] Juraj Wiedermann. Parallel Turing machines [J]. Techn. Rep. RUU-CS-84-11, Dept. of Computer Science,University of Utrecht, Utrecht.1984:1-35.
[96]邬江兴，罗兴国.基于认知的主动重构计算体系[M].上海:中国工程院高端论坛论文集，2012
[97] Arthur S. Bland, Ricky A. Kendall, Douglas B. Kotheet al. Jaguar: The world’s most powerful computer[J]. Memory (TB),2009,300（62）:362.
[98] Kevin J. Barker, Kei Davis, Adolfy Hoisieet al. Entering the petaflop era: the architecture andperformance of Roadrunner[M].IEEE Press，2008:1.
[99]蒋本珊.计算机组成原理[M]:清华大学出版社，2004
[100]王一宾，韩承双.计算机体系结构与软件体系结构[J].安徽建筑工业学院学报(自然科学版),2005,1:20.
[101]黄海，吴昊.基于CUDA的立体匹配算法评估工具研究[J].系统仿真学报,2012,24（001）:154-157.
[102] Cristiana Amza, Alan L. Cox, Sandhya Dwarkadaset al. Treadmarks: Shared memory computing onnetworks of workstations [J]. Computer,1996,29（2）:18-28.
[103] Reiner Hartenstein. A decade of reconfigurable computing: a visionary retrospective [M]. IEEE Press，2001:642-649.
[104] David Kielpinski, Chris Monroe,David J. Wineland. Architecture for a large-scale ion-trap quantumcomputer [J]. Nature Publishing Group,2002:709-711.
[105] Forest Baskett, Alan Jay Smith. Interference in multiprocessor computer systems with interleavedmemory [J]. Communications of the ACM,1976,19（6）:327-334.
[106] Tim J. Todman, George A. Constantinides, Steven JE Wiltonet al. Reconfigurable computing:architectures and design methods [M].IET，2005:193-207.
[107] Malcolm Ware, Karthick Rajamani, Michael Floydet al. Architecting for power management: TheIBM POWER7approach [M]. IEEE，2010:1-11.
[108] B. Sinharoy, R. Kalla, W. J. Starkeet al. IBM POWER7multicore server processor [J]. IBM Journal ofResearch and Development,2011,55（3）:1.
[109] Victor W. Lee, Changkyu Kim, Jatin Chhuganiet al. Debunking the100X GPU vs. CPU myth: anevaluation of throughput computing on CPU and GPU[M].ACM，2010:451-460.
[110] Johannes Gilger, Johannes Barnickel, Ulrike Meyer. GPU-Acceleration of Block Ciphers in theOpenSSL Cryptographic Library [J]. Information Security,2012:338-353.
[111] InfiniBand Trade Association. InfiniBand Architecture Specification: Release1.0[M]. InfiniBand TradeAssociation，2000.
[112] Jiuxing Liu, Jiesheng Wu, Dhabaleswar K. Panda. High performance RDMA-based MPI implementationover InfiniBand [J]. International Journal of Parallel Programming,2004,32（3）:167-198.
[113] Eugene L. Lawler, Jan Karel Lenstra, AHG Rinnooy Kanet al. The traveling salesman problem: a guidedtour of combinatorial optimization[M].Wiley Chichester，1985.
[114]李弋.云环境中典型应用的I/O优化策略研究[D].上海:复旦大学，2012
[115] Peter M. Kogge. EXECUBE-a new architecture for scaleable MPPs[M]IEEE，1994:77-84
[116] David Patterson, Thomas Anderson,Neal Cardwellet al. A case for intelligent RAM[J]. Micro,IEEE,1997,17（2）:34-44.
[117] Ashish A. Pandya. Complex symbol evaluation for programmable intelligent search memory[M].GooglePatents，2010.
[118] Todd M. Austin. DIVA: A reliable substrate for deep submicron microarchitecture design [M]. IEEE，1999:196-207.
[119] Seung-Moon Yoo, Jose Renau, Michael Huanget al. Flexram architecture design parameters [J]. Centerfor Supercomputing Research and Development (CSRD), Tech, Rep.2000:1584.
[120] Mark Oskin, Frederic T. Chong, Timothy Sherwood. Active pages: A computation model for intelligentmemory [M]: IEEE Computer Society，1998
[121] Jack J. Dongarra, David W. Walker. The quest for petascale computing[J]. Computing in Science&Engineering,2001,3（3）:32-39.
[122] Steve Leibson, James Kim. Configurable processors: a new era in chip design [J]. Computer,2005,38（7）:51-59.
[123] Peter M. Chen, Edward K. Lee,Garth A. Gibsonet al. RAID: High-performance, reliable secondarystorage[J]. ACM Computing Surveys (CSUR),1994,26（2）:145-185.
[124] Jim Gray. Tape is dead, disk is tape, flash is disk, RAM locality is king[EB/OL].:Gong ShowPresentation at CIDR,2006-12-10
[125] Erik Riedel, Garth Gibson, Christos Faloutsos. Active storage for large-scale data mining andmultimedia applications[M].Citeseer，1998:62-73.
[126] Parthasarathy Ranganathan. From microprocessors to nanostores: Rethinking system building blocks forthe data-centric era [J],2011:1.
[127]Kevin Lim,Parthasarathy Ranganathan,Jichuan Changet al. Understanding and des igning new serverarchitectures for emerging warehouse-computing environments [M]. IEEE，2008:315-326
[128] Kevin Lim, Parthasarathy Ranganathan, Jichuan Changet al. Understanding and designing new serverarchitectures for emerging warehouse-computing environments [M]. IEEE，2008:315-326.
[129] David G. Andersen, Jason Franklin, Michael Kaminskyet al. FAWN: A fast array of wimpy nodes [M].ACM，2009:1-14.
[130] Behrouz A. Forouzan. TCP/IP protocol suite [M]. McGraw-Hill, Inc.，2002.
[131]Shen Gao Jianliang Xu, Bingsheng He, Byron Choi Haibo Hu. PCMLogging: Reducing TransactionLogging Overhead with PCM [EB/OL].2011:https://www.comp.hkbu.edu.hk/~xujl/Papers/cikm2011_gao.pdf
[132] Cristiano P. Costa, Italo S. Cunha, Alex Borgeset al. Analyzing client interactivity in streaming media
[M]. ACM，2004:534-543.
[133] Lei Guo, Songqing Chen, Zhen Xiaoet al. DISC: Dynamic interleaved segment caching for interactivestreaming [M]. IEEE，2005:763-772.
[134]王国权，李弋，叶德建.针对IPTV快进快退优化的媒体封装格式[J]. Computer Engineering,2010,36（6）:218-220.