From: Riccardo Veraldi Subject: asymmetric performance for network I/O Date: Tue, 06 Oct 2009 15:23:09 +0200 Message-ID: <4ACB44BD.5010706@cnaf.infn.it> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit To: kvm@vger.kernel.org Return-path: Received: from iris.cnaf.infn.it ([131.154.3.7]:46574 "EHLO iris.cnaf.infn.it" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757292AbZJFNqf (ORCPT ); Tue, 6 Oct 2009 09:46:35 -0400 Received: from darwin.cnaf.infn.it (darwin.cnaf.infn.it [131.154.3.242]) by iris.cnaf.infn.it (8.14.2/8.14.2) with ESMTP id n96DNhQ9025480 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK) for ; Tue, 6 Oct 2009 15:23:43 +0200 Sender: kvm-owner@vger.kernel.org List-ID: Hello, I did some tsts with KVM virtual machines. I gan good results on gigabit interface with inbound traffic, I get almost gigabit saturation (1Gb/s). I have lower perforance outbound that is around 500Mbit/s. Is it normal in your opinion ? I mean is there some hard coded limit for network I/O performance ? any experience of someone else about this topic ? thanks Rick