From mboxrd@z Thu Jan 1 00:00:00 1970 From: Anthony Liguori Subject: Re: kvm live migration reliable? Date: Fri, 31 Jul 2009 08:11:44 -0500 Message-ID: <4A72ED90.8060004@codemonkey.ws> References: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org To: Kent Tong Return-path: Received: from mail-gx0-f213.google.com ([209.85.217.213]:37394 "EHLO mail-gx0-f213.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752228AbZGaNLr (ORCPT ); Fri, 31 Jul 2009 09:11:47 -0400 Received: by gxk9 with SMTP id 9so3681518gxk.13 for ; Fri, 31 Jul 2009 06:11:46 -0700 (PDT) In-Reply-To: Sender: kvm-owner@vger.kernel.org List-ID: Kent Tong wrote: > Hi, > > I was wondering the kvm live migration function is reliable? I read that if > the TSC is not monotonically increasing, then the guest will hang. How to > prevent this? > KVM transfers the tsc as part of the CPU state so it will continue to increase monotonically. However, if you're migrating between two CPUs of different frequency, the result will be that the TSC frequency will change. This could have negative effects depends on how much the guest relies on the TSC. The worst case would be if your guest was Linux and was using TSC as a time source. You would likely see severe time drift. A PV clock source (like KVM_CLOCK) helps this quite a bit. Regards, Anthony Liguori > Thanks! > > > -- > To unsubscribe from this list: send the line "unsubscribe kvm" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html >