摘要
随着网络技术的发展,更多的商业应用服务开始由IP网络承载,这对网络的可靠性和可用性提出了更高的要求。IP协议存在一定的鲁棒性和生存性,但故障恢复依赖收敛时间,往往达不到关键业务和实时流量的要求。多协议标签交换技术MPLS具备故障快速响应能力,提供包括重路由和保护切换在内的故障快速恢复机制。快速重路由结合了重路由和保护切换的优点,具有更好的性能和适用性。本文深刻研究了快速重路由机制的支撑技术,主要介绍了Makam模型和Haskin模型两种常用的快速重路由算法。
本文首先通过模拟实验验证快速重路由的性能优势。恢复性能与网络状态参数有密切关系,本文从Makam模型和Haskin模型入手分析了快速重路由性能与网络的关系,并对两个模型进行了实验模拟,对比其优缺点,验证分析推论,为算法设计提供指导。
针对当前网络面临的多故障环境,课题考虑设计一种高效的多故障恢复算法。在深入分析当前多故障恢复技术的前提下,我们认为对网络采取合适的划分,将多故障恢复分解为局部故障恢复是非常有效的方法。在此基础上,本文提出了多约束条件下的分域恢复思想,基于多约束条件将网络划分为多SPD,实现SPD内的故障恢复。在该思想的指导下,本文设计了基于多约束条件的分域恢复算法MCSR,将恢复时间和共享资源作为约束条件划分SPD,在域内采用Haskin模型进行故障恢复。
本文最后对MCSR算法进行了实验模拟,设置了多个多故障场景测试其恢复性能。实验结果表明MCSR算法在对多故障进行恢复时能有效降低恢复时间,减少报文丢失,提高资源利用率,达到了高效多故障恢复的目标。通过对算法的QoSP性能分析,对比Haskin模型和MCSR的QoSP值,表明MCSR算法不仅能快速恢复故障,而且能满足QoS的要求。
With the development of network technology, IP network carries more economic applications than ever, which requires the increasment of reliability and availability. The IP protocols have definite robustness and survivability, but the recovery speed relies on the convergence time, which cannot satisfy the constraints of some critical serveices and real-time traffic. Multi-protocol label switching has embedded the fast response ability for the network faults, meanwhile providing faults recovery mechanisms including rerouting and protection switching. Fast rerouting combines the advantages of rerouting and protection switching, so it has better performance and applicability. The article deeply reserches on the supporting technology of fast rerouting, and mainly introduces Makam and Haskin model.
We validate the performance of fast rerouting through simulating experiments. The recovery performance has consanguineous relationship with network statues. We analyze this relationship through Makam and Haskin models, and simulate the two models in order to compare the advantages and disadvantages, as well as validate our deduations to guide the algorithms design.
As the network has been faced with the multi-fault environment, we consider designing effective multi-fault recovery algorithm. Based on the sufficient analysis of the current multi-fault recovery technologies, we decide that splitting to convert the problem to local recovery is an effective method. So we propose ana idea of segmented recovery under multi-constraint, and design an algorithm named MCSR which takes the recovery time and shared resources as the constraints. This algorithm segmented the network into several SPD and adopts the Haskin model as the in-domain recovery.
We simulate MCSR under several scenarios to show the performance.The result indicates that the proposed algorithm can reduce the recovery time and packet loss efficiently, while enhancing the utilization rate of resources. It reaches the requirements of multi-fault recovery efficiency. Through the QoSP analyzing, the article compares the performance between Haskin’s model and the MCSR algorithm. The comparing results proves that the proposed algorithm fulfil the fast fault recovery as well as the QoSP requirement.
引文
[1] Latest Global Internet Statistical Information . 2007, http://www.apira.org/html/report/report_00_en.htm
[2] Introduction to High Availability Networking, http://www.cisco.com/networkers/nw00/pres/2609_6-28.pdf
[3] Thomas M. Chen, Tae H. Oh, Reliable services in MPLS. IEEE Communications Magazine, 1999.12: 58-62
[4] Changcheng Huang, Ken Owens, Srinivas Makam, Building reliable MPLS networks using a path protection mechanism. IEEE Communications Magazine, 2002.3: 156-162
[5] V. Sharma, F. Hellstrand, Framework for Multi-Protocol Label Switching (MPLS)-based Recovery. RFC3469, 2003.2
[6] S.Makam, V.Sharma, K.Owens, C.Huang, Protection/Restoration of MPLS Networks. , 1999.10
[7] Dimitry Haskin, Ram Krishnan, Robert Boyd, Alan Hannan, A Method for Setting an Alternative Label Switched Paths to Handle Fast Reroute. < draft-haskin-mpls-fast-reroute-00.txt>, 1999.6
[8] E. Rosen, A. Viswanathan, R. Callon, Multiprotocol Label Switching Architecture. RFC3031, 2001.1
[9] L. Andersson, P. Doolan, N. Feldman, A. Fredette, B. Thomas, LDP Specification. RFC3036, 2001.1
[10] P. Pan, G. Swallow, A. Atlas, Fast Reroute Extensions to RSVP-TE for LSP Tunnels, RFC4090. 2005.5
[11] D.Adami, C.Callegari, D.Ceccarelli, S.Giordano, M.Pagano, Design and development of MPLS-based recovery strategies in NS2. IEEE GLOBECOM proceedings, 2006
[12] Santos Kumar Das, P. Venkataram, A method of designing a path restoration scheme for MPLS based network. IEEE Conference Paper, 2002.5: 218-222
[13] Lemma Hundessa, Jordi Domingo-Pascual, Fast rerouting mechanism for a protected label switched path. IEEE Communications Magazine, 2001.3: 527-530
[14] Lemma Hundessa, Jordi Domingo-Pascual, Reliable and fast rerouting mechanism for a protected label switched path. Computer Communications and Networks Proceeding, 2002.3: 1608-1612
[15] Sung-Heon Shin, Hae-Joon Shin, Young-Talc Kim, Packet lossless fast rerouting scheme with minimized buffer delay. Network Operations and Management Symposium, 2004.7: 219-231
[16] Sangsik Yoon, Hyunseok Lee, Deokjai Choi, Youngcheol Kim, Gueesang Lee,Mike Lee, An efficient recovery mechanism for MPLS-based protection LSP. ICATM, 2001.7: 75-79
[17]赵丹,胡晓峰,吴纯青. MPLS故障恢复机制仿真与分析.第十三届青年通信学术会议, 2008.10(已录用)
[18] The Network Simulator-NS2, http://nsnam.isi.edu/nsnam/index.php/
[19] Gaeil Ahn, Woojik Chun, Design and implementation of MPLS network simulator (MNS) supporting QoS. Information Networking, 2001.7: 694-699
[20] The RSVP-TE Network Simulator Home Page: http://netgroupserv.iet.unipi.it/rsvp-te_ns/
[21] The OSPF-TE Network Simulator Home Page: http://netgroupserv.iet.unipi.it/ospf-te_ns/
[22] Lemma Hundessa, Jordi Domingo-Pascnal, Optimal and Guaranteed Alternative LSP for Multiple Failures. ICCCN, 2004.8: 59-64
[23]李彬,陈向东.一种MPLS网络的故障恢复方案.微电子学与计算机, 2007,24(1): 126-129
[24] Daniel W. Hong, Choong Seon Hong, Woo-Sung Kim, A Segment-based Protection Scheme for MPLS Network Survivability. Network Operations and Management Symposium, 2006.7
[25] Jose L Marzo, Eusebi Calle, Caterina Scoglio, Tricha Anjali, Adding QoS protection in order to enhance MPLS QoS routing. IEEE International Conference, 2003.4: 1973-1977
[26] F. Le Faucheur, W. Lai, Requirements for Support of Differentiated Services-aware. RFC3564, 2003.7
[27] F. Le Faucheur, W. Lai, Requirements for Support of Differentiated Services-aware. RFC3564, 2003.7
[28] Alex Raj, Oliver C. Ibe, A survey of IP and multiprotocol label switching fast reroute schemes, Computer Networks. 2007(51): 1882–1907
[29] Christopher Metz, IP Protection and Restoration. IEEE Internet Computing, 2000.3
[30]郎静,王佳坤,张连芳. MPLS网络中一种可扩展的动态恢复机制.微计算机信息, 2007,23(4): 183-185
[31] Anjali Agarwal, Rohan Deshmukh. Ingress failure recovery mechanisms in MPLS network. MILCOM Proceedings, 2000.1: 1150-1153
[32] LDP FRR技术白皮书, Huawei Technologies Co., Ltd, 2007
[33] MPLS TE FRR技术白皮书, Huawei Technologies Co., Ltd, 2007
[34]康剑,曹阳,葛非,王悦伟.一种新的MPLS网络故障恢复算法.计算机应用研究, 2005(8): 207-209
[35]徐迎晓,张根度.一种MPLS故障恢复算法.计算机研究与发展, 2003.4(4):596-601
[36] Wisitsak Sa-Ngiamsak, Ruttikorn Varakulsiripunth. Local reverse backup and resiliency buffer approaches for MPLS-based recovery. TENCON, 2004.8: 609-612
[37] Luyuan Fang, Alia Atlas, Fabio Chiussi, Kireeti Kompella, George Swallow. LDP Failure Detection and Recovery. IEEE Communications Magazine, 2004.10: 117-123
[38] Radim Bartos, Mythilikanth Raman, Arun Gandhi. New Approaches to Service Restoration in MPLS-based Networks. EUROCON, 2001.2: 58-61
[39]李晓东. MPLS技术与实现.电子工业出版社, 2002.11
[40] Luc De Ghein著,陈麟帆译, MPLS技术构架.人民邮电出版社, 2008.1
[41] Jim Guichard, Franciois Le Faucheur, Jean-Philippe Vasseur著,陈武译, MPLS网络设计权威指南.人民邮电出版社, 2007.1
[42]徐雷鸣,庞博,赵耀. NS与网络模拟.人民邮电出版社, 2003.11
