A performance study of a Network Front End

ABSTRACT

A Network Front End (NFE) is a mini-computer which is used to connect a host computer to the communicating network. The purpose of using an NFE, as contrasted to connecting the host computer directly to the network, is to reduce the processing load imparted to the host computer by the network interfacing software. This paper presents a comparative evaluation of the NFE attachment method in relation to the host direct connection method. The comparative evaluation addressed two major areas: the impact on the host of the network interface software and end-to-end network performance. Various evaluation experiments were conducted, the results of which indicate that the NFE is a promising alternative to host direct connection in terms of performance in the WWMCCS Intercomputer Network (WIN). In addition, the NFE also has the capability to satisfy the standardized high capacity host-AUTODIN II interface requirement and to provide other features, such as direct terminal to network connection and network security, not available in the WIN architecture. It is recommended that refinement of the NFE and further development be continued.

REFERENCES

• 1: CCTC, NFE Development Plan, CCTC/DCA, 3 September 1976.
• 2: Bruckner, S. M., Network Front End Evaluation Plan, MTR-5292, MITRE Washington, July 1977.
• 3: Poh, S. S., Stoneburner, P. D., and Wood, D. C., Network Front End Evaluation Report, MITRE Washington, August 1978.
• 4: CSC, Host to Front End Processor Module Functional Description, 21 February 1977.
• 5: CSC, Host to Front End Processor Protocol Interface (HFPI) Functional Description, 21 February 1977.
• 6: Center for Advanced Computation, Networking Research in Front Ending and Intelligent Terminals, Experimental Network Front End Functional Description, CAC Document Number 221, CAC, University of Illinois, 15 January 1977.
• 7: Center for Advanced Computation, Network Research in Front Ending and Intelligent Terminals, Experimental Network Front End Software Functional Description, CAC Document Number 223, CAC, University of Illinois, 1 July 1977.
• 8: CCTC, Network Front-End Software User's Manual, CCTC/DCA, 3 April 1978.
• 9: Smith, D. C., Synthetic Jobstreams for the H6000, WP-19470, MITRE Washington, 2 Jun 1978.
• 10: Bruckner, S. M., Performance Measurement and Analysis Using the Trace Tool, MTR-5240, MITRE Washington, May 1975.
• 11: Bruckner, S. M., Communications Measurement and Analysis for WWMCCS ADP, MTR-5256, MITRE Washington, November 1975.
• 12: Janczewski, M. G., Network Performance Measurement in the H6000—Using the Trace Facility, WP-19482, MITRE Washington, May 1978.
• 13: CCTC, Memory Utilization Monitor,/CPU Monitor/and Tape Monitor—User's Guide, C702/CCTC/DCA, 1 April 1977.
• 14: Poh, S. S., Benvenuto, A. A., and Eagle, J. E., End-to-End Throughput Analysis in a Packet-Switched Network, Proceedings of the ACM-NBS Fifteenth Annual Technical Symposium, 17 June 1976, pp. 149-155.
• 15: Wood, D. C., Measurement of User Traffic Characteristics on ARPANET, Proc. 4th Data Communications Symposium, October 1975.

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Performance attributes

Additional Classification:
  C. Computer Systems Organization
  C.3 SPECIAL-PURPOSE AND APPLICATION-BASED SYSTEMS

General Terms:
Experimentation, Performance