Hi, assuming that the data rate was the same for all the experiments, with higher packet sizes the number of packets to be processed at the switch per time unit decreases. Hence, (assuming that processing capabilities at the switch is the bottleneck, and NOT forwarding capabilities) more data can be send per time unit which is why queue lengths are shorter which in turn is why latency decreases as well.
Best, Philip > Am 18.08.2016 um 19:07 schrieb Nam Bong Ha <over3...@gmail.com>: > > Hi, all > > I read the following thesis. > > http://archive.openflow.org/downloads/technicalreports/MS_Thesis_Polito_2009_Manuel_Palacin_OpenFlow.pdf > > At the figure 4.22, 4.23,4.24 and etc, I could find latency getting lower > when packet length becomes bigger. > > What is the reason of this? > > ThankYou! > _______________________________________________ > openflow-discuss mailing list > openflow-discuss@lists.stanford.edu > https://mailman.stanford.edu/mailman/listinfo/openflow-discuss
_______________________________________________ openflow-discuss mailing list openflow-discuss@lists.stanford.edu https://mailman.stanford.edu/mailman/listinfo/openflow-discuss
