[Cake] measuring "active flows"
Dave Taht
dave.taht at gmail.com
Fri Apr 8 13:09:30 EDT 2016
One of the nifty things about cake is that we are in a position to get
more of a grip on "active" flows with lots of contention. I have done
a bit of this ad-hoc here and there, with 100s of flows, the number
"active" (with a queue) at any given RTT are much less (at least with
drop rather than ecn)
https://www.researchgate.net/profile/S_Oueslati/publication/234778751_Evaluating_the_number_of_active_flows_in_a_scheduler_realizing_fair_statistical_bandwidth_sharing/links/5451fab10cf2bf864cbab666.pdf
(If you can't tell, I'm gradually trying to collect papers that were
influential on me)
Pretty much everything else luca and jim roberts has written is pure gold, too.
https://scholar.google.com/citations?user=Q36qUAsAAAAJ&hl=en
--
Dave Täht
Let's go make home routers and wifi faster! With better software!
https://www.gofundme.com/savewifi
More information about the Cake
mailing list