From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: [bug] stuck localhost TCP connections, v2.6.26-rc3+ Date: Fri, 30 May 2008 20:18:39 +0200 Message-ID: <20080530181839.GA31915@elte.hu> References: <20080526115628.GA31316@elte.hu> <20080529084524.GA24892@elte.hu> <20080529112257.GA18130@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: LKML , Netdev , "David S. Miller" , "Rafael J. Wysocki" , Andrew Morton To: Ilpo J?rvinen Return-path: Received: from mx2.mail.elte.hu ([157.181.151.9]:47716 "EHLO mx2.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752403AbYE3STE (ORCPT ); Fri, 30 May 2008 14:19:04 -0400 Content-Disposition: inline In-Reply-To: <20080529112257.GA18130@elte.hu> Sender: netdev-owner@vger.kernel.org List-ID: * Ingo Molnar wrote: > after about 50 bootups i got a hung test again: > > titan:~/tip> netstat -nt > Active Internet connections (w/o servers) > Proto Recv-Q Send-Q Local Address Foreign Address > State > tcp 0 0 10.0.1.14:22 10.0.1.16:58062 ESTABLISHED > tcp 0 0 10.0.1.14:22 10.0.1.16:60109 ESTABLISHED > tcp 0 86368 10.0.1.14:43914 10.0.1.16:3632 ESTABLISHED > > and this time with CUBIC_TCP disabled - so that was a red herring. ah, in retrospect i realized that this test had one flaw: some of the systems i the build cluster already ran a newer kernel and hence were targets for this bug. so i turned off CONFIG_TCP_CONG_CUBIC on all the testboxes and rebooted the cluster boxes into 2.6.25, and the hung sockets are now gone. (about 150 successful iterations) i did another change as well: i removed the localhost distcc component. I'll reinstate that now to make sure it's really related to TCP_CONG_CUBIC and not to localhost networking. Ingo