QoS management is switch-dependent. Some switches (like a reference Stanford openflow switch and derived from them) use queue facilities, another (for ex. HP HW switches) drives by PCP/TOS bits -- look at docs & sources. For ex. -- at lib/netdev.c of reference Stanford openflow switch you can see some useful comments about queue management implementation with standard Linux traffic control machinery.
16/02/2013 11:03 -0800, Geetha S wrote: > My goal is to introduce QoS in openflow networks. > > > I will looking at the Slicing tutorial on Openflow.org and came across > an experiment that deals with queues and its min rate etc.I am using > Mininet to create hosts and switch and POX controller. I am a beginner > towards POX and have understood l2 forwarding switch behavior. > > > I am currently interested in working with queues and take it from > there. I came across this link and i was stuck at the add-queue part > (dint really understand what happened after that), > > https://mailman.stanford.edu/pipermail/mininet-discuss/2010-November/000172.html > > > I am not sure if i am proceeding in the right direction. Is slicing > really required for me to deal with QoS ? > > > Please guide me. > > > Thanks, > Geetha > _______________________________________________ > 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
