From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Jarosch Subject: Re: [PATCH] tcp FRTO: in-order-only "TCP proxy" fragility workaround Date: Wed, 13 Aug 2008 10:00:28 +0200 Message-ID: <200808131000.29788.thomas.jarosch@intra2net.com> References: <200808120946.19459.thomas.jarosch@intra2net.com> <20080812.011822.155249871.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Cc: ilpo.jarvinen@helsinki.fi, billfink@mindspring.com, fragabr@gmail.com, netdev@vger.kernel.org, kaber@trash.net, sr@securenet.de, netfilter-devel@vger.kernel.org, kadlec@blackhole.kfki.hu To: David Miller Return-path: In-Reply-To: <20080812.011822.155249871.davem@davemloft.net> Content-Disposition: inline Sender: netfilter-devel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Tuesday, 12. August 2008 10:18:22 David Miller wrote: > From: Thomas Jarosch > Date: Tue, 12 Aug 2008 09:46:17 +0200 > > > David, I agree with you, though I'm not sure about the end user > > experience: > > We had the same situation with ECN and window scaling, and my proposal > is the same as how we handled those situations involving broken > middleware boxes. Yes, that is true. IMHO there's a slight difference with FRTO trouble compared to ECN/window scaling issues: ECN trouble -> No access at all Broken window scaling -> Large transfers don't work MTU issues -> No access at all / large transfers don't work FRTO problems -> Hard to spot as they only happen when packet loss occurs. Though I guess Ilpo knows best if there's an "easy" way to detect this or not. Thomas