From owner-freebsd-net Sun Jul 18 0:57:11 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 9D00114BE0 for ; Sun, 18 Jul 1999 00:57:07 -0700 (PDT) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id HAA17810; Sun, 18 Jul 1999 07:27:05 +0200 From: Luigi Rizzo Message-Id: <199907180527.HAA17810@labinfo.iet.unipi.it> Subject: Re: dummynet -> rate limiting To: wollman@khavrinen.lcs.mit.edu (Garrett Wollman) Date: Sun, 18 Jul 1999 07:27:05 +0200 (MET DST) Cc: nagao@iij.ad.jp, des@flood.ping.uio.no, net@FreeBSD.ORG In-Reply-To: <199907172241.SAA24085@khavrinen.lcs.mit.edu> from "Garrett Wollman" at Jul 17, 99 06:40:49 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1426 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > Ok, i have no objections in principle, but i fail to see the use of > > pps limiting. What does it model in a context (IP) where packets for sure > > are not constant size ? > > Surely I shouldn't need to give this lesson to you, in particular, > Luigi. As we all know, performance of network elements can be broken > down into two components: per-packet cost, and per-bit (mostly ... > Cisco added a packet-rate-limiting feature in their ISP train some > time ago, and it made it into 12.0 on certain platforms, so at least > one big Cisco customer must think it's useful. But maybe IOS does not have bit-rate-limiting ? With dummynet you can achieve almost the same effect by setting a pipe with a very low bit rate, and yes, this will penalize big packets but if the purpose is protecting yourself, being conservative is not a such concern. Plus, for good protection you'd want both pps and bit-rate limiting, so this raises the issue whether these two features on a pipe should be mutually exclusive or not (in practice the first is ok since you can cascade pipes...). What i was wondering is whether this pps fits some economic or network model where a packet is major cost (or billing) component. other than that, i really don't mind having 'pps' limitations, or other things, in dummynet -- i am just trying to see if they solve problems that cannot be handled with already existing features. cheers luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message