From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755433Ab2JPH7I (ORCPT ); Tue, 16 Oct 2012 03:59:08 -0400 Received: from tee.schottelius.org ([77.109.138.222]:34614 "EHLO mx3.schottelius.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753532Ab2JPH7H (ORCPT ); Tue, 16 Oct 2012 03:59:07 -0400 X-Greylist: delayed 608 seconds by postgrey-1.27 at vger.kernel.org; Tue, 16 Oct 2012 03:59:07 EDT Date: Tue, 16 Oct 2012 09:51:22 +0200 From: Nico Schottelius To: LKML Subject: Very bad performance on Qemu/KVM Message-ID: <20121016075122.GA25824@schottelius.org> Mail-Followup-To: Nico Schottelius , LKML MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: echo $message | gpg -e $sender -s | netcat mailhost 25 X-Netzseite: http://www.nico.schottelius.org/ X-System-Info: brief (Linux 3.5.6-1-ARCH x86_64) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Good morning LKML, I see extremely slow virtual machines running on qemu-kvm-0.12.1.2-2.295.el6_3.2.x86_64 on Centos 6.3 with Linux 2.6.32-279.5.2.el6.x86_64 on a 64 core machine using AMD Opteron(TM) Processor 6276 (dell r815). The guests are Centos 6.3, Linux 2.6.32-220.13.1.el6.x86_64. The disk images reside on local storage, ext4 filesystem. I checked that kvm is loaded and enabled. I changed -drive to use if=virtio, -net to use model=virtio. >>From the processes running it seems that cpu performance is the bottleneck. I added -cpu host to skip the emulation, but this did not fix the issue. Comparing the performance to qemu running on a xenserver, performance is at most half of it. Does anyone of you have a good pointer how to debug this further or are there any known issues with the given setup? I'm happy for every pointer I get, as there seems to be no easy to spot issue. Cheers, Nico -- PGP key: 7ED9 F7D3 6B10 81D7 0EC5 5C09 D7DC C8E4 3187 7DF0