whenever I get some time (and the right tcp patches) is fiddle with two l4s approaches in fq_codel. The easy one is to leverage ce_threshold, accepting 1ms of queuing delay.
A better one, I think is to leverage the sparse flows optimization in fq_codel to always mark when it's an old flow, and never mark when it's new. That appears to be the most scalable, down to us or less. Both ideas would take some work to do in cake, and I lack time and the dctcp patches to try it. -- Dave Täht CEO, TekLibre, LLC http://www.teklibre.com Tel: 1-669-226-2619 _______________________________________________ Cake mailing list [email protected] https://lists.bufferbloat.net/listinfo/cake
