From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrey Korolyov Subject: QEMU P2P migration speed Date: Tue, 04 Feb 2014 21:06:36 +0400 Message-ID: <52F11E1C.8060608@xdel.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit To: kvm@vger.kernel.org Return-path: Received: from mail-la0-f49.google.com ([209.85.215.49]:48629 "EHLO mail-la0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752125AbaBDRGm (ORCPT ); Tue, 4 Feb 2014 12:06:42 -0500 Received: by mail-la0-f49.google.com with SMTP id y1so6553559lam.8 for ; Tue, 04 Feb 2014 09:06:40 -0800 (PST) Received: from [192.168.10.117] ([213.33.189.50]) by mx.google.com with ESMTPSA id gi5sm25907394lbc.4.2014.02.04.09.06.38 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Tue, 04 Feb 2014 09:06:39 -0800 (PST) Sender: kvm-owner@vger.kernel.org List-ID: Hello, I`ve got strange results during benchmarking migration speed for different kinds of loads on source/target host: when source host is 'empty', migration takes approx. 30 percent longer than the same for host already occupied by one VM with CPU overcommit ratio=1. [src host, three equal vms, each with ability to eat all cores once] [tgt host, one VM, with same appetite and limitations] All VMs was put into cgroups with same cpu ceiling and cpu shares values. Migration time is almost independent of VM RSS(varies by ten percent at maximum), for situation when VM is active on target host, time is about 85 seconds to migrate 8G between hosts, and when it is turned off, migration time *increasing* to 120s. For curious ones, frequency management is completely inactive on both nodes, neither CStates mechanism. Interconnection is relatively fast (20+Gbit/s by IPoIB). Anyone have a suggestions on how to possibly explain/fix this?