From owner-freebsd-current Sat Apr 25 15:09:36 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA27274 for freebsd-current-outgoing; Sat, 25 Apr 1998 15:09:36 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA27206 for ; Sat, 25 Apr 1998 15:09:32 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.8.8/8.8.8) id SAA16900; Sat, 25 Apr 1998 18:08:16 -0400 (EDT) (envelope-from wollman) Date: Sat, 25 Apr 1998 18:08:16 -0400 (EDT) From: Garrett Wollman Message-Id: <199804252208.SAA16900@khavrinen.lcs.mit.edu> To: "John S. Dyson" Cc: trost@cloud.rain.com (Bill Trost), current@FreeBSD.ORG Subject: Re: Bandwidth throttling etc. In-Reply-To: <199804252056.PAA12181@dyson.iquest.net> References: <908.893476488@cloud.rain.com> <199804252056.PAA12181@dyson.iquest.net> Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG < said: > AFAIK, and I am NOT a networking export, we need to improve the sockets > layer as much as the lower level networking code. Actually, we need to get rid of the socket layer altogether. It was a good idea to save coding effort and .text space when all we had were VAXen with 1 MB of core, but if we really want network protocols to go fast, we need to write all of the upper-layer functionality in ways which are specific to the needs of the network protocols (and, eventually, to the applications). I have done about 10% of this work for UDP, but suspended that effort last fall to work on more immediate concerns. (If you want to look at how this might be taken to extremes, look at the Exokernel, . I don't think we'll see production systems go this far because of serious manageability issues, but it's one very compelling demonstration of that end of the spectrum.) Anyone who's interested in discussing this more, please contact me privately; perhaps we can put together a mailing-list. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message