![]() |
![]() ![]() ![]() |
![]() |
![]() |
|||||||||||||||||||||||||
![]() |
||||||||||||||||||||||||||||
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
![]() |
![]() |
![]() |
|||||||||||||||||||||||||
![]() ![]() ![]() ![]() ![]() |
Constant Bitrate Background As we are conducting tests on real equipment and not on simulator, it it often difficult to perform tests under different network conditions. We could use something like dummynet to similute packet loss and delay, however, we would not be able to achieve the capacities nor the latencies required for our studies. In order to generate cbr background traffic, we run another instance of iperf, on a separate pair of machines along the same link. We ensure that the two streams share a bottleneck. Usage A single udp cbr stream is produced by running iperf between two hosts. Then the tcp stack under investigation is run after about 5 seconds. This is to allow the cbr flow to stablise before running the test. The background udp stream rate is varied from 25 to line rate in steps of 25mbit/sec, and the results logged. Download
Results MB-NG
DataTAG
Summary TODO
|
|||||||||||||||||||||||||||
![]() |
![]() |
![]() |
||||||||||||||||||||||||||
![]() |
![]() |
![]() |
||||||||||||||||||||||||||
© 2001-2003, Yee-Ting Li, email: ytl@hep.ucl.ac.uk,
Tel: +44 (0) 20 7679 1376, Fax: +44 (0) 20 7679 7145 Room D14, High Energy Particle Physics, Dept. of Physics & Astronomy, UCL, Gower St, London, WC1E 6BT |
||||||||||||||||||||||||||||
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |