Date: Sat, 21 Jul 2007 11:03:42 +0300 From: Artyom Viklenko <artem@aws-net.org.ua> To: Artem Belevich <fbsdlist@src.cx> Cc: freebsd-net@freebsd.org, Julian Elischer <julian@elischer.org> Subject: Re: 6.2 mtu now limits size of incomming packet Message-ID: <46A1BDDE.5080403@aws-net.org.ua> In-Reply-To: <ed91d4a80707201534q6e32a9c5j8803ec96fa7f6515@mail.gmail.com> References: <200707150237.l6F2bAgZ011098@redrock.karels.net> <469E0FFF.8070802@seclark.us> <20070720172021.8EA3D13C4B3@mx1.freebsd.org> <46A10063.9010902@elischer.org> <46A10860.50804@es.net> <ed91d4a80707201534q6e32a9c5j8803ec96fa7f6515@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Artem Belevich wrote: > Here's one example where MTU!=MRU would be useful. > > Think of asymmetric bandwith-limited ADSL links. Lower MTU would allow > lower TX latency for high priority packets when upstream is saturated, > yet large MRU on the downstream would be great for downloads. > > Right now with 6.2 one has to trade off lower latency for faster download. > > --Artem You can prioritize small packets with ACKs, for example, by other techniques - ALTQ one of them. Unconditional lovering MTU even on ADSL tend to loss throughtput. And let's think about TCP MSS. When TCP connection establishes, TCP stack uses MTU as measure to choose MSS. Any two hosts, connected to single Layer2 network MUST use same MTU. Any other cases lead to hard-to-solve problems. This is all IMHO. But I would not like to see different MTU and MRU on my Ethernet interfaces! :) -- Sincerely yours, Artyom Viklenko. ------------------------------------------------------- artem@aws-net.org.ua | http://www.aws-net.org.ua/~artem FreeBSD: The Power to Serve - http://www.freebsd.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?46A1BDDE.5080403>