
›› 2018,Vol. 33 ›› Issue (2): 400416.doi: 10.1007/s1139001818263
所属专题： Computer Architecture and Systems； Computer Networks and Distributed Computing
• Theory and Algorithms • 上一篇 下一篇
Xi Wang^{1,2}, Member, CCF, JianXi Fan^{1*}, Member, CCF, ChengKuan Lin^{1}, Member, CCF, JingYa Zhou^{1}, Member, CCF, Zhao Liu^{1}
Xi Wang^{1,2}, Member, CCF, JianXi Fan^{1*}, Member, CCF, ChengKuan Lin^{1}, Member, CCF, JingYa Zhou^{1}, Member, CCF, Zhao Liu^{1}
数据中心网络的性能在很大程度上决定云计算的性能，但随着应用需求的不断提高，数据中心网络中的服务器数量变得越来越庞大。如何将大量服务器连接起来，从而构建一个性能良好的数据中心网络，是提升云计算性能所面临的一个挑战。传统的树型数据中心网络存在带宽瓶颈和单点失效等问题，目前提出的DCell、BCube和FiConn等数据中心网络具有较大的带宽和容错性，但DCell和FiConn在交换机失效时服务器间的容错路径长度较大；而BCube在规模较大时对交换机性能有较高要求。综合上述考虑，基于具有优良性能的交叉立方体，我们提出了一种新的以服务器为中心的数据中心网络，称为BCDC。进一步，我们研究了BCDC网络的顶点度数，通信算法以及容错路由算法。另外，我们分析了BCDC上路由算法的性能及时间复杂度并进行相应的模拟实验。该研究将为新型数据中心网络的设计和实现提供重要依据。
[1] Harris D. Ballmer's millionserver claim doesn't seem so crazy. https://gigaom.com/2013/07/17/ballmersmillionserverclaimdoesntseemsocrazy/#comments, July 2013. [2] Dignan L. AWS financials on deck:The road to 3 million servers in operation. http://www.zdnet.com/article/awsfinancialsondecktheroadto3millionserversinoperation/, April 2015. [3] AlFares M, Loukissas A, Vahdat A. A scalable, commodity data center network architecture. In Proc. the ACM SIGCOMM Conf. Data Communication, August 2008, pp.6374. [4] Guo C X, Wu H T, Tan K, Shi L, Zhang Y G, Lu S W. DCell:A scalable and faulttolerant network structure for data centers. In Proc. the ACM SIGCOMM Conf. Data Communication, August 2008, pp.7586. [5] Li D, Guo C X, Wu H T, Tan K, Zhang Y G, Lu S W. FiConn:Using backup port for server interconnection in data centers. In Proc. IEEE INFOCOM, April 2009, pp.22762285. [6] Guo C X, Lu G H, Li D, Wu H T, Zhang X, Shi Y F, Tian C, Zhang Y G, Lu S W. BCube:A high performance, servercentric network architecture for modular data centers. In Proc. the ACM SIGCOMM Conf. Data Communication, August 2009, pp.6374. [7] Greenberg A, Hamilton J R, Jain N, Kandula S, Kim C, Lahiri P, Maltz D A, Patel P, Sengupta S. VL2:A scalable and flexible data center network. In Proc. the ACM SIGCOMM Conf. Data Communication, August 2009, pp.5162. [8] AbuLibdeh H, Costa P, Rowstron A, O'Shea G, Donnelly A. Symbiotic routing in future data centers. In Proc. ACM SIGCOMM, Aug.30Sept.3, 2010, pp.5162. [9] Yu Y, Qian C. Space shuffle:A scalable, flexible, and highperformance data center network. IEEE Trans. Parallel and Distributed Systems, 2016, 27(11):33513365. [10] Zheng K, Wang L, Yang B H, Sun Y, Uhlig S. LazyCtrl:A scalable hybrid network control plane design for cloud data centers. IEEE Trans. Parallel and Distributed Systems, 2017, 28(1):115127. [11] Bhuyan L N, Agrawal D P. Generalized hypercube and hyperbus structures for a computer network. IEEE Trans. Computers, 1984, C33(4):323333. [12] Leiserson C E. Fattrees:Universal networks for hardwareefficient supercomputing. IEEE Trans. Computers, 1985, 34(10):892901. [13] Dally W J. Performance analysis of kary ncube interconnection networks. IEEE Trans. Computers, 1990, 39(6):775785. [14] Xiang D, Zhang Y L, Pan Y. Practical deadlockfree faulttolerant routing in meshes based on the planar network fault model. IEEE Trans. Computers, 2009, 58(5):620633. [15] Xiang D. Deadlockfree adaptive routing in meshes with faulttolerance ability based on channel overlapping. IEEE Trans. Dependable and Secure Computing, 2011, 8(1):7488. [16] Lin D, Liu Y, Hamdi M, Muppala J. FlatNet:Towards a flatter data center network. In Proc. IEEE Global Communications Conf., December 2012, pp.24992504. [17] Wang T, Su Z Y, Xia Y, Qin B, Hamdi M. NovaCube:A low latency Torusbased network architecture for data centers. In Proc. IEEE Global Communications Conf., December 2014, pp.22522257. [18] Wang T, Su Z Y, Xia Y, Liu Y, Muppala J, Hamdi M. SprintNet:A high performance servercentric network architecture for data centers. In Proc. IEEE Int. Conf. Communications, June 2014, pp.40054010. [19] Wang T, Su Z Y, Xia Y, Muppala J, Hamdi M. Designing efficient high performance servercentric data center network architecture. Computer Networks, 2015, 79:283296. [20] Wang T, Su Z Y, Xia Y, Hamdi M. CLOT:A costeffective lowlatency overlaid Torusbased network architecture for data centers. In Proc. IEEE Int. Conf. Communications, June 2015, pp.54795484. [21] Li D W, Wu J, Liu Z Y, Zhang F. Towards the tradeoffs in designing data center network architectures. IEEE Trans. Parallel and Distributed Systems, 2017, 28(1):260273. [22] Efe K. A variation on the hypercube with lower diameter. IEEE Trans. Computers, 1991, 40(11):13121316. [23] Cull P, Larson S M. The Möbius cubes. IEEE Trans. Computers, 1995, 44(5):647659. [24] Abraham S, Padmanabhan K. The twisted cube topology for multiprocessors:A study in network asymmetry. Journal of Parallel and Distributed Computing, 1991, 13(1):104110. [25] Fan J X, He L Q. BC interconnection networks and their properties. Chinese Journal of Computers, 2003, 26(1):8490. (in Chinese) [26] Wang D J. Hamiltonian embedding in crossed cubes with failed links. IEEE Trans. Parallel and Distributed Systems, 2012, 23(11):21172124. [27] Kulasinghe P, Bettayeb S. Embedding binary trees into crossed cubes. IEEE Trans. Computers, 1995, 44(7):923929. [28] Fan J, Lin X, Jia X. Optimal path embedding in crossed cubes. IEEE Trans. Parallel and Distributed Systems, 2005, 16(12):11901200. [29] Efe K. The crossed cube architecture for parallel computation. IEEE Trans. Parallel and Distributed Systems, 1992, 3(5):513524. [30] Chang C P, Sung T Y, Hsu L H. Edge congestion and topological properties of crossed cubes. IEEE Trans. Parallel and Distributed Systems, 2000, 11(1):6480. [31] Efe K, Blackwell P K, Slough W, Shiau T. Topological properties of the crossed cube architecture. Parallel Computing, 1994, 20(12):17631775. [32] Kulasinghe P D. Connectivity of the crossed cube. Information Processing Letters, 1997, 61(4):221226. [33] Fan J X, Jia X H. Edgepancyclicity and pathembeddability of bijective connection graphs. Information Sciences, 2008, 178(2):340351. [34] Yang X F, Dong Q, Tang Y Y. Embedding meshes/tori in faulty crossed cubes. Information Processing Letters, 2010, 110(14/15):559564. [35] Zhou S M. The conditional diagnosability of crossed cubes under the comparison model. International Journal of Computer Mathematics, 2010, 87(15):33873396. [36] Dong Q, Zhou J L, Fu Y, Yang X F. Embedding a mesh of trees in the crossed cube. Information Processing Letters, 2012, 112(14/15):599603. [37] Cheng B L, Fan J X, Jia X H, Zhang S K. Independent spanning trees in crossed cubes. Information Sciences, 2013, 233:276289. [38] Cheng B L, Fan J X, Jia X H, Wang J. Dimensionadjacent trees and parallel construction of independent spanning trees on crossed cubes. Journal of Parallel and Distributed Computing, 2013, 73(5):641652. [39] Chen H C, Kung T L, Hsu L Y. 2disjointpathcoverable panconnectedness of crossed cubes. The Journal of Supercomputing, 2015, 71(7):27672782. [40] Chen H C, Zou Y H, Wang Y L, Pai K J. A note on path embedding in crossed cubes with faulty vertices. Information Processing Letters, 2017, 121:3438. [41] Cheng B L, Wang D J, Fan J X. Constructing completely independent spanning trees in crossed cubes. Discrete Applied Mathematics, 2017, 219:100109. [42] Diestel R. Graph Theory (4th edition). Springer, 2010. [43] Ghemawat S, Gobioff H, Leung S T. The Google file system. In Proc. the 19th ACM Symp. Operating Systems Principles, October 2003, pp.2943. [44] Dean J, Ghemawat S. MapReduce:Simplified data processing on large clusters. Communications of the ACM, 2008, 51(1):107113. 
No related articles found! 

版权所有 © 《计算机科学技术学报》编辑部 本系统由北京玛格泰克科技发展有限公司设计开发 技术支持：support@magtech.com.cn 总访问量： 