From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: many sockets, slow sendto Date: Thu, 22 Mar 2007 00:26:42 +0100 Message-ID: <4601BF32.5070300@cosmosbay.com> References: <20070306182039.GJ25760@galon.ev-en.org> <45FF185B.4070007@fw.hu> <20070319.161611.70218081.davem@davemloft.net> <4600592E.80605@fw.hu> <460064C6.5030302@cosmosbay.com> <4601A949.2000005@fw.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-2; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: David Miller , baruch@ev-en.org, netdev@vger.kernel.org To: Zacco Return-path: Received: from www.cosmosbay.com ([86.65.150.131]:58881 "EHLO gw1.cosmosbay.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1751480AbXCUX1N (ORCPT ); Wed, 21 Mar 2007 19:27:13 -0400 In-Reply-To: <4601A949.2000005@fw.hu> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Zacco a =E9crit : > So, my worry is confirmed then. But how could that delay disappear wh= en=20 > splitting the sender and receiver on distinct hosts? Even in that cas= e=20 > the good socket must be found somehow. When the receiver and sender are on the same machine, the sendto() pass= the=20 packet to loopback and enters the receiving side. With that many socket= s, the=20 time to go through all sockets maybe 100 us. So your sendto() seems to = be=20 slow, but the slow part is the receiver. If you put two machines, the sender might send XX.XXX frames per second= (full=20 speed), but the receiver might handle 5% of them and drop 95% This is all speculation, since you didnt gave us the exact setup you us= e.