From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?ISO-8859-1?Q?D=E2niel?= Fraga Subject: Re: [PATCH] tcp FRTO: in-order-only "TCP proxy" fragility workaround Date: Thu, 28 Aug 2008 18:49:19 -0300 Message-ID: <20080828184919.611dd578@tux> References: <20080819213417.45133573@tux> <20080822183224.2d52f16c@tux> <20080822.143709.65615512.davem@davemloft.net> <20080823111446.06a350a2@tux> <20080824163843.33b4f890@tux> <20080826141812.589848a0@tux> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: David Miller , thomas.jarosch@intra2net.com, billfink@mindspring.com, Netdev , Patrick Hardy , netfilter-devel@vger.kernel.org, kadlec@blackhole.kfki.hu To: "Ilpo =?ISO-8859-1?Q?J=E4rvinen?=" Return-path: Received: from an-out-0708.google.com ([209.85.132.249]:26070 "EHLO an-out-0708.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754923AbYH1Vta (ORCPT ); Thu, 28 Aug 2008 17:49:30 -0400 Received: by an-out-0708.google.com with SMTP id d40so101511and.103 for ; Thu, 28 Aug 2008 14:49:28 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Tue, 26 Aug 2008 23:40:58 +0300 (EEST) "Ilpo J=E4rvinen" wrote: > while [ : ]; do > date "+%s.%N" > cat /proc/net/{netstat,snmp,tcp} > sleep 1 > done Ok; Let's try again, now with more data (I hope): 1) tcpdump (just port 995): http://www.abusar.org/stall/dump-client http://www.abusar.org/stall/dump-server http://www.abusar.org/stall/dump-server-loopback I don't know if loopback is useful, just in case... 2) the above script : http://www.abusar.org/stall/script-client-log.txt http://www.abusar.org/stall/script-server-log.txt 3) and strace from the client Claws Mail: http://www.abusar.org/stall/strace-client-claws-mail.txt I forgot to use the -r option in strace. But when Claws-mail stalls, it gives the following multiple times: read(4, 0xf48704, 4096) =3D -1 EAGAIN (Resource tempora= rily unavailable) select(5, [4], [4], NULL, NULL) =3D 1 (out [4]) writev(4, [{"xxxxxxxxxxxxxxx"..., 108}], 1) =3D 108 select(5, [4], [], NULL, NULL) =3D 1 (in [4]) read(4, "xxxxxxxxxxxxxxxxxxxxxx"..., 4096) =3D 108 Well, I hope this time you have more information and I hope I didn't f= orget anything.=20 If not, let's keep trying. Important: these data were collected with frto disabled (0) and htb di= sabled too. So it isn't related to frto, neither htb. Thank you! --=20