The Performance of Software Multicast-Reflector Implementations for Multi-player Online Games

  • Daniel Bauer
  • Sean Rooney
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2816)


Massive multi-player online games are large distributed applications where thousands of participants exchange data. Existing solutions based on central servers face scalability problems. We study a hybrid solution between the peer-to-peer and central server models that divides a large game into several federated small games. The central component of this architecture is a multicast reflector. We present two efficient software implementations that have been developed as Linux kernel extensions and compare them with our user-space implementation. The comparison is based on performance measurements done on actual implementations.


Multicast Group Device Driver Large Game Memory Page Address List 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Abdelkhalek, A., Bilas, A., Moshovos, A.: Behavior and performance of interactive multi-player game servers. In: Proceedings of the International IEEE Symposium on the Performance Analysis of Systems and Software, ISPASS-2001 (November 2001)Google Scholar
  2. 2.
    Bauer, D., Iliadis, I., Rooney, S., Scotton, P.: Communication Architectures for Massive Multi-Player Games. Technical Report RZ3500, IBM Research (June 2003),
  3. 3.
    Bharambe, A., Rio, S., Seshan, S.: Mercury: A Scalable Publish-Subscribe System for Internet Games. In: NetGames 2002 – First Workshop on Network and System Support for Games, Braunschweig, Germany (April 2002)Google Scholar
  4. 4.
    Clendenin, M.: Dual-channel DDR chips aim for 5.3 Gbyte/s bandwith. EETimes (August 2002),
  5. 5.
    Cox, A.: Network buffers and memory management. Linux Journal (September 1996)Google Scholar
  6. 6.
    Morse, K.: Interest management in large-scale distributed simulations. Tech. report 96-27, Dept. of Information and Computer Science, University of California, Irvine (1996) Google Scholar
  7. 7.
    OpenSkies. OpenSkies Network Architecture,
  8. 8.
    OpenSkies. Openskies performance test: Demonstration of scalability,
  9. 9.
    Pettersson, M.: Linux x86 Performance-Monitoring Counters Driver (March 2003),
  10. 10.
    Postel, J.: User datagram protocol. Request for Comments 768, Internet Engineering Task Force (August 1980)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Daniel Bauer
    • 1
  • Sean Rooney
    • 1
  1. 1.Zurich LaboratoryIBM ResearchRüschlikonSwitzerland

Personalised recommendations