Advertisement

A Highly-Efficient Approach to Adaptive Load Balance for Scalable TBGP

  • Lei Gao
  • Mingche Lai
  • Kefei Wang
  • Zhengbin Pang
Conference paper
Part of the Communications in Computer and Information Science book series (CCIS, volume 396)

Abstract

Multi-threading technique used in BGP protocol empowers the remarkable performance enhancement, but load imbalance on different threads could also become the bottleneck of performance and scalability especially on occurrence of route instability such as route outburst update. To resolve this issue, an effective adaptive route load partition (ARLP) algorithm for TBGP is proposed to provide load balance on different threads and alleviate the influence of route instability. The ARLP algorithm dynamically dispatches sessions on different threads by the statistical number of route update from each neighbor session, and distributes sessions with high load over different threads, to maximize the computing resources and ensure the minimum load difference on threads. Experimental results on dual quad-core Xeon sever show that ARLP algorithm could achieve steady load balance ratio ranging from 0.11 to 0.14 under four thread configurations. We also derive MUTPT as the metric to evaluate the route update performance of TBGP, and the MUTPT samples with ARLP are averagely reduced by 46.2%, 51.8%, 63.2% and 70.3% compared to those with static load-distribution method, yielding good performance improvement and session scalability for TBGP.

Keywords

BGP multi-threaded load balance performance scalability 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Huston G.: The BGP Report for 2005(EB/OL), http://www.potaroo.net/papers/isoc/2006-06/bgpupds.html, 2006-06-01/2007-05-01
  2. 2.
    Nguyen, K.-K., Jaumard, B.: A scalable and distributed architecture for bgp in next generation routers. In: Proc. of ICC Communications Workshops, pp. 1–5 (2009)Google Scholar
  3. 3.
    Taft, N.: The basics of bgp routing and its performance in today’s internet. In: Proc. of RHDM High-speed Networks and Multimedia Workshop, France, pp. 35–42 (2001)Google Scholar
  4. 4.
    Wang, K., Gao, L., et al.: A scalable multithreaded BGP architecture for next generation router. In: EMC 2011, vol. (102), pp. 395–404 (2011)Google Scholar
  5. 5.
    Xu, K., He, H.: BGP parallel computing model based on the iteration tree. Journal of China Universities of Posts and Telcommunications 15(suppl.), 1–8 (2008)CrossRefzbMATHGoogle Scholar
  6. 6.
    Liu, Y., Zhang, S., Wang, B.: MR-PBGP: A multi-root tree model for parallel bgp. In: Proc. of High Performance Computing and Communications, pp. 1211–1218 (2012)Google Scholar
  7. 7.
    Zhang, X., Lu, X., Su, J., et al.: SDBGP: A Scalable, Distributed BGP Routing Protocol Implementation. In: Proc. of the 12th IEEE International Conference on High Performance Switching and Routing, pp. 191–196 (2011)Google Scholar
  8. 8.
    Griffin, T., Shepherd, F.B., Wilfong, G.: The stable paths problem and interdomain routing. IEEE/ACM Transactions on Networking 10(1), 232–234 (2002)CrossRefGoogle Scholar
  9. 9.
    Varadhan, K., Govindan, R., Estrin, D.: Persistent route oscillations in inter-domain routing. Computer Networks 32(1), 1–16 (2000)CrossRefGoogle Scholar
  10. 10.
    Liu, X., Xiao, L.: Inbound traffic load balancing in BGP multi-homed stub networks. In: Proc. of International Conference on Distributed Computing Systems, pp. 369–376 (2008)Google Scholar
  11. 11.
    Tong, R., Zhu, X.: A Load Balancing Strategy Based on the Combination of Static and Dynamic. In: Proc. of Database Technology and Applications, pp. 1–4 (2011)Google Scholar
  12. 12.
    Guo, F., Chen, J., Li, W., Chiueh, T.: Experiences in Building A Multihoming Load Balancing System. In: Proceedings of IEEE INFOCOM, pp. 1241–1251 (2004)Google Scholar
  13. 13.
    Fujinoki, H.: Improving Reliability for Multi-Home Inbound Traffic: MHLB/I Packet- Level Inter-Domain Load-Balancing. In: Proc. of International Conference on Availability, Reliability and Security, pp. 248–256 (2009)Google Scholar
  14. 14.
    Zhou, S., Chen, J., et al.: Approximate load balance based on id/locator split routing architecture. In: World Congress on Information and Communication Technologies, pp. 981–986 (2012)Google Scholar
  15. 15.
    Smiljanić, A.: Rate and Delay Guarantees Provided by Clos Packet Switches with Load Balancing. IEEE Trans. Netw., 170–181 (February 2008)Google Scholar
  16. 16.
    Antić, M., Smiljanić, A.: Oblivious Routing Scheme Using Load Balancing Over Shortest Paths. In: Proc. ICC 2008 (2008)Google Scholar
  17. 17.
    Antić, M., Smiljanić, A.: Routing with Load Balancing: Increasing the Guaranteed Node Traffics. IEEE Commun. Lett., 450–452 (June 2009)Google Scholar
  18. 18.
    Kostic-Ljubisavljevic, A., Radonjic, V., et al.: Load Balance Routing for Interconnected Next Generation Networks Performances Improvement, pp.42–46 (2011)Google Scholar
  19. 19.
    Kodialam, M., Lakshman, T.V., Sengupta, S.: Traffic-Oblivious Routing for Guaranteed Bandwidth Performance. IEEE Commun. Mag. 45(4), 46–51 (2007)CrossRefGoogle Scholar
  20. 20.
    Antić, M., Maksic, N., et al.: Two phase load balanced routing using OSPF. IEEE Journal of Selected Areas in Communications 28(1), 51–59 (2010)Google Scholar
  21. 21.
    Lei, G., Lai, M., Gong, Z.: A practical non-blocking route propagation technology for threaded BGP. In: ICSCC 2010, pp. 206–211 (2010)Google Scholar
  22. 22.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Lei Gao
    • 1
  • Mingche Lai
    • 1
  • Kefei Wang
    • 1
  • Zhengbin Pang
    • 1
  1. 1.College of ComputerNational University of Defense TechnologyChangshaChina

Personalised recommendations