From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: TCP congestion control for fast, short-distance networks ? Date: Wed, 4 Apr 2007 10:39:44 -0700 Message-ID: <20070404103944.1b15de16@freekitty> References: <20070403091656.769e5368@localhost.localdomain> <200704032113.l33LDVhp025035@expredir6.cites.uiuc.edu> <20070404080743.5e67245a@localhost.localdomain> <8C944E41FEE84B1-244-BFF@FWM-D41.sysops.aol.com> <20070404153911.GA32019@galon.ev-en.org> <8C944E930D99B5D-244-E4B@FWM-D41.sysops.aol.com> <20070404161121.GB32019@galon.ev-en.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: davef1624@aol.com, netdev@vger.kernel.org To: Baruch Even Return-path: Received: from smtp.osdl.org ([65.172.181.24]:49198 "EHLO smtp.osdl.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752201AbXDDRjv (ORCPT ); Wed, 4 Apr 2007 13:39:51 -0400 In-Reply-To: <20070404161121.GB32019@galon.ev-en.org> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Wed, 4 Apr 2007 19:11:21 +0300 Baruch Even wrote: > * davef1624@aol.com [070404 19:03]: > > Thanks - so you are suggesting we enable 802.3 flow-control / pause-frames? > > (it's currently disabled) > > I do, but do test it before you bet on it. I've never tested such a > scenario but from my experience the lower the rtt the lesser are the > problems that the high speed algorithms are trying to solve. > > Baruch > > > > > -----Original Message----- > > From: baruch@ev-en.org > > To: davef1624@aol.com > > Cc: shemminger@linux-foundation.org; netdev@vger.kernel.org > > Sent: Wed, 4 Apr 2007 8:39 AM > > Subject: Re: TCP congestion control for fast, short-distance networks ? > > > > * davef1624@aol.com [070404 18:29]: > > >Hello, > > > > > >We are currently using both 1 Gb & 10 Gb links, that interconnect > > several > > servers that are very *local* to > > >each other. > > >Typical RTT times range from 0.2 ms - 0.3 ms. > > > > > >We are currently using TCP reno - is there a more suitable congestion > > control > > algorithm for our > > >application, > > >especially using the 10 Gb links ? > > >(Most of the High-Speed TCP algorithms seem suitable for large RTT, > > long-distance networks). > > > > I'm not aware of any tests for high speed links with very low RTTs, but > > I suspect that the new algorithms will not change much, if the > > connections you have are indeed local than the Ethernet pause mechanism > > is more effective for the flow control you need. > > > > Baruch Pause helps (if the hardware works). But on very high speed links with low RTT, your main source of throttling is lack of bus bandwidth on receiver. -- Stephen Hemminger