Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

Performance Analysis of Queuing Algorithms on Datakit T1 Trunk Lines

Michael J. Hawley

EECS Department
University of California, Berkeley
Technical Report No. UCB/CSD-89-547
December 1989

http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/CSD-89-547.pdf

Datakit VCS is a networking system which integrates local and wide area networks, and attempts to provide good interactive response in the presence of bulk transfer traffic. The queueing algorithms used at the trunk interface are the chief means of providing good interactive response. This report compares the performance of fifo queueing and per-channel queueing by implementing these disciplines on a programmable trunk board and measuring the throughput and interactive delay. These measurements show that (1) fifo queueing delays messages up to ten times longer than per-channel queueing when the trunk is heavily loaded, and that (2) per-channel queueing does not adversely affect trunk throughput or delay when the trunk is not loaded.


BibTeX citation:

@techreport{Hawley:CSD-89-547,
    Author = {Hawley, Michael J.},
    Title = {Performance Analysis of Queuing Algorithms on Datakit T1 Trunk Lines},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {1989},
    Month = {Dec},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/5902.html},
    Number = {UCB/CSD-89-547},
    Abstract = {Datakit VCS is a networking system which integrates local and wide area networks, and attempts to provide good interactive response in the presence of bulk transfer traffic. The queueing algorithms used at the trunk interface are the chief means of providing good interactive response. This report compares the performance of fifo queueing and per-channel queueing by implementing these disciplines on a programmable trunk board and measuring the throughput and interactive delay. These measurements show that (1) fifo queueing delays messages up to ten times longer than per-channel queueing when the trunk is heavily loaded, and that (2) per-channel queueing does not adversely affect trunk throughput or delay when the trunk is not loaded.}
}

EndNote citation:

%0 Report
%A Hawley, Michael J.
%T Performance Analysis of Queuing Algorithms on Datakit T1 Trunk Lines
%I EECS Department, University of California, Berkeley
%D 1989
%@ UCB/CSD-89-547
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/5902.html
%F Hawley:CSD-89-547