Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262460AbUKVXmC (ORCPT ); Mon, 22 Nov 2004 18:42:02 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262459AbUKVXlq (ORCPT ); Mon, 22 Nov 2004 18:41:46 -0500 Received: from advect.atmos.washington.edu ([128.95.89.50]:59628 "EHLO advect.atmos.washington.edu") by vger.kernel.org with ESMTP id S262445AbUKVXiZ (ORCPT ); Mon, 22 Nov 2004 18:38:25 -0500 Message-ID: <41A27868.80703@atmos.washington.edu> Date: Mon, 22 Nov 2004 15:38:16 -0800 From: Harry Edmon User-Agent: Mozilla Thunderbird 0.8 (X11/20040926) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Martin Josefsson CC: Con Kolivas , Stephen Hemminger , linux-kernel@vger.kernel.org Subject: Re: Network slowdown from 2.6.7 to 2.6.9 References: <419A9151.2000508@atmos.washington.edu> <20041116163257.0e63031d@zqx3.pdx.osdl.net> <419BA5C4.4020503@atmos.washington.edu> <1100722571.20185.9.camel@tux.rsn.bth.se> <419BBF57.3040808@atmos.washington.edu> <1100727847.20185.31.camel@tux.rsn.bth.se> In-Reply-To: <1100727847.20185.31.camel@tux.rsn.bth.se> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: -12.68 () AWL,BAYES_00 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1348 Lines: 50 Tried them all - none of them helped. Use "ntop" I can see that my throughput on the Intel gigabit ethernet interface on the system maxes out at 15.2 Mbps with 2.6.9. With 2.6.7 it made it to 35 Mbps. Does anyone have any other suggestions as to what to look for to diagnose this problem? Martin Josefsson wrote: >On Wed, 2004-11-17 at 22:15, Harry Edmon wrote: > > >>No, it is not loaded. Here is the list of loaded modules under 2.6.9: >> >> > >Ok, then it's something else, maybe the TSO changes... > >Here's some thigs to try: > >You can see the current settings, TSO etc, with > >'ethtool -k eth0' >and disable TSO with >'ethtool -K eth0 tso off' > >If that doesn't help there's more things to try. > >I don't know anything about the diffrent TCP congestion algorithms >(Stephen does), but you could try disabling BIC. > >echo 0 > /proc/sys/net/ipv4/tcp_bic > >(continuing with things I hardly know anything about...) > >Or maybe the autotuning receive buffer, try disabling that with > >echo 0 > /proc/sys/net/ipv4/tcp_moderate_rcvbuf > > >Stephen, any more things to try ? > > > - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/