Multimedia Tools and Applications

, Volume 2, Issue 3, pp 253–271 | Cite as

Factors influencing ATM adapter throughput

  • Andrew Rindos
  • Steven Woolet
  • David Cosby
  • Leonard Hango
  • Mladen Vouk
Article

Abstract

Analytical and experimental evaluations show that ATM adapters can perform quite close to their designed limits, provided that they are used in a properly configured environment with series resources capable of sustaining the desired throughputs. While the media speed may be 155, 100 or 25 Mbps, there are a number of factors that will determine the final achieved maximum throughput observed by a user of a ATM adapter. One is the overhead inherent to ATM, such as the 5 byte ATM header that accompanies every 48 bytes of data sent. Others are inherent in the protocol used in the communication layers above ATM (e.g., pacing and retransmission associated with TCP/IP). Still others are dependent upon the processor speed and operating system used by the adapter host system. This paper discusses those parameters that, based on our experience, can have a considerable impact on the throughput of an ATM adapter. Turboways1 25, 100 and 155 ATM adapter measurements illustrate these issues. These represent end-to-end (application layer-to-application layer) throughput measurements, involving all supporting hardware (workstations, 8260 ATM hub/switch, 8282 ATM concentrator, etc.) and all protocol layers, operating system, etc.

Key words

ATM Networking Performance Multimedia 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    ATM Forum. LAN Emulation Over ATM Specification-Version 1.0, 1995. LAN Emulation SWG Drafting Group.Google Scholar
  2. 2.
    Laura A.Chappell and Dan E.Hakes, Novell's Guide to NetWare LAN Analysis. Novell Press, San Jose, CA, USA, second edition, 1994.Google Scholar
  3. 3.
    DouglasComer. Internetworking with TCP/IP Volume 1: Principles, protocols, and architecture. Prentice Hall, Englewood Cliffs, NJ. USA, 2nd edition, 1991.Google Scholar
  4. 4.
    H. M. Deitel and M. S. Kogan. The Design of OS/2. Addison-Wesley Publishing Co., Inc., 1992.Google Scholar
  5. 5.
    W.E.Denzel, A.P.J.Engbersen, and I.Iliadis, A flexible shared-buffer switch for ATM at Gb/s rates. Computer Networks and ISDN Systems, 27:611–624, 1995.Google Scholar
  6. 6.
    Hewlett-Packard Company. Netperf: A Network Performance Benchmark, 1995. Revision 2.0.Google Scholar
  7. 7.
    IBM. LAN Technical Reference: IEEE 802.2 and NETBIOS Application Program Interfaces, December 1993. SC30-3587.Google Scholar
  8. 8.
    IBM Specification Sheet. ATM Campus Networking Products, 1995. G325-3505.Google Scholar
  9. 9.
    ISO/IEC 8802-3 [ANSI/IEEE Std 802.3]. CSMA/CD Access method and physical layer specifications, 1993.Google Scholar
  10. 10.
    ISO/IEC 8802-5 [ANSI/IEEE Std 802.5] Token ring access method and physical layer specifications, 1992.Google Scholar
  11. 11.
    V. Jacobson, R. Braden, and D. Borman. TCP extensions for high performance. Network Working Group, RFC-1323, May 1992.Google Scholar
  12. 12.
    M. Laubach. Classical IP and ARP over ATM. Internet RFC-1577, January 1994.Google Scholar
  13. 13.
    S. J. Mastrianni, Writing OS/2 2.1 Device Drivers. Van Nostrand Reinhold, 1993.Google Scholar
  14. 14.
    R. Onvural. Asynchronous Transfer Mode Networks: Performance Issues. Artech House, 1995.Google Scholar
  15. 15.
    D. E. Reich. Designing OS/2 Applications. John Wiley & Sons, Inc., 1993.Google Scholar
  16. 16.
    A.J. Rindos, S.P. Woolet, D. W. Cosby, R. Ram, L.P. Hango, and M. A. Vouk. Some factors influencing ATM adapter throughput. In B. Furht and M.H. Hamza, editors, Proc. of the Second IASTED/ISMM Int'l. Conf. Distributed Multimedia Systems and Applications, papes 15–18, Stanford, CA USA, August 1995.Google Scholar
  17. 17.
    D. Saha, D. Kandlur, T. Barzilai, Z. Shae, and M. Willebeek-LeMair. A video conferencing testbed on ATM: Design, implementation and optimization. In Proc. Int. Conf. on Multimedia Communications and Systems, pages 23–30, 1995.Google Scholar
  18. 18.
    TomShanley and DonAnderson. PCI System Architecture (rev 2.0 compliant), volume 4 of PC system architecture series. MindShare Press, Richardson, TX, USA, 1993.Google Scholar
  19. 19.
    M. Vouk and A. Rindos. An evaluation of ATM technology. In Proc. IASTED/ISMM Intl. Conf on Distributed Multimedia Sys. and Appl., Honolulu, Hawaii, August 1994.Google Scholar

Copyright information

© Kluwer Academic Publishers 1996

Authors and Affiliations

  • Andrew Rindos
    • 1
  • Steven Woolet
    • 1
  • David Cosby
    • 1
  • Leonard Hango
    • 1
  • Mladen Vouk
    • 2
  1. 1.Networking Hardware DivisionIBM CorporationResearch Triangle ParkUSA
  2. 2.Department of Computer ScienceNorth Carolina State UniversityRaleighUSA

Personalised recommendations