Date: Tue, 03 Oct 2000 11:29:01 -0700 From: Julian Elischer <julian@elischer.org> To: Garrett Wollman <wollman@khavrinen.lcs.mit.edu>, net@freebsd.org Subject: Re: SACK in FreeBSD TCP. Message-ID: <39DA256D.8077E698@elischer.org> References: <39DA1CEE.ADCEA934@elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Garrett Wollman wrote: > > <<On Tue, 03 Oct 2000 09:07:34 -0700, Julian Elischer <julian@elischer.org> said: > > > In the past I had something to do with some systems that used > > "vegas-like" > > The bug in Vegas was that it broke congestion control. > > -GAWollman check out the first article on this page .. http://www.cs.inf.ethz.ch/~bolliger/ for an interesting analysis.... Unfortunatly New-reno was not in the comparison, but your suggestion that Vegas broke congestion control is not supported. Another more relevant reference is the LINUX Vegas page: http://www.cs.washington.edu/homes/cardwell/linux-vegas/ They have quite a few results there... Most interesting is the comment that SACK can also produce the same gain, and that SACK+Vegas gives littel extra gain. However for Non-SACK clients, Vegas can give good results without relying on the client to upgrade.. Also they point out that Vegas is a better net-citizen in that it tends to use up less buffers in routers in the cloud. Anyway It looks like there are definitly lots of improvements we can still make.. -- __--_|\ Julian Elischer / \ julian@elischer.org ( OZ ) World tour 2000 ---> X_.---._/ presently in: Perth v To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?39DA256D.8077E698>