From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: Network hangs with 2.6.30.5 Date: Fri, 02 Oct 2009 14:38:30 +0200 Message-ID: <4AC5F446.4090202@gmail.com> References: <20090907072143.GA5966@ff.dom.local> <20091001.154913.88345178.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: David Miller , jarkao2@gmail.com, holger.hoffstaette@googlemail.com, Netdev , Evgeniy Polyakov To: =?ISO-8859-15?Q?Ilpo_J=E4rvinen?= Return-path: Received: from gw1.cosmosbay.com ([212.99.114.194]:56480 "EHLO gw1.cosmosbay.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756634AbZJBMig (ORCPT ); Fri, 2 Oct 2009 08:38:36 -0400 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Ilpo J=E4rvinen a =E9crit : > On Fri, 2 Oct 2009, Ilpo J=E4rvinen wrote: >=20 >> On Thu, 1 Oct 2009, David Miller wrote: >> >>> From: Jarek Poplawski >>> Date: Mon, 7 Sep 2009 07:21:43 +0000 >>> >>>> While Eric is analyzing your data, I guess you could try reverting >>>> some stuff around this tcp_tw_recycle, and my tcp ignorance would >>>> point these commits for the beginning: >>>> >>>> http://git.kernel.org/?p=3Dlinux/kernel/git/stable/linux-2.6.30.y.= git;a=3Dcommitdiff;h=3Dfc1ad92dfc4e363a055053746552cdb445ba5c57 >>>> http://git.kernel.org/?p=3Dlinux/kernel/git/stable/linux-2.6.30.y.= git;a=3Dcommitdiff;h=3Dc887e6d2d9aee56ee7c9f2af4cec3a5efdcc4c72 >>> Ilpo's cleanup (the second commit listed) looks most likely to >>> be a possibility. >>> >>> But I surely cannot find any bugs in it, even after studying it >>> a few times. >>> >>> Ilpo could you audit it one more time for us just in case? >> Argh, not that one ...the jungle of negations. But I'll try to go it= =20 >> through once more but I tell you I did go through those negations mu= ltiple=20 >> times already before submitting it :-). >> >>> I also looked through all the TCP commits in 2.6.29 to 2.6.30 >>> and I could not find anything else that might cause stalls with >>> time-wait recycled connections. >> What about the more than 64k connections change a9d8f9110d7e953c2f2 = (or=20 >> its fixes), it might be another possibility? ...It certainly does=20 >> something related to reuse and happens to be in the correct time fra= me...=20 >> (I've added Evgeniy). I scratched my head to reproduce the conditions of hang but failed. I am pretty sure both commits are OK (yours and mine), maybe a brute fo= rce git bisection is needed.