Date: Fri, 20 Jun 2014 12:19:14 +1000 From: grenville armitage <garmitage@swin.edu.au> To: freebsd-net@freebsd.org Subject: Re: LEDBAT (RFC-6817)i n FreeBSD as mod_cc(9)? Message-ID: <53A39A22.5090802@swin.edu.au> In-Reply-To: <1819611615.20140617124843@serebryakov.spb.ru> References: <1863738251.20140616172546@serebryakov.spb.ru> <CALCpEUHSsPifQ3PPgEmT5TipOoEQ=YREK6bJb6UPUX2zHDyySg@mail.gmail.com> <1819611615.20140617124843@serebryakov.spb.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
On 06/17/2014 18:48, Lev Serebryakov wrote: > Hello, hiren. > You wrote 17 =D0=B8=D1=8E=D0=BD=D1=8F 2014 =D0=B3., 11:41:35: > >>> It looks like, that some TCP connections could benefit from LEDBAT= >>> (RFC-6871) cognestion control algorithm (not all, of course, it sho= uld not >>> be default). I wonder what you think of using cc_cdg ? We did some initial exploratio= n of using cc_cdg (v0.1) in LEDBAT-like contexts in http://caia.swin.edu.= au/cv/garmitage/things/CameraReady-CDGscaveneger-LCN2013-preprint.pdf) [..] > The problem is it seems that "someone" needs to extend set of hooks = in > mod_cc substantially. It is more than "write one more module" and, IM= HO, > should be discussed with net-related people :) I'm also curious about implementing LEDBAT -- what sort of functionality = do you sense is missing from current mod_cc and h_ertt ? cheers, gja
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?53A39A22.5090802>