From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:46509) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XjQMw-0005VD-EO for qemu-devel@nongnu.org; Wed, 29 Oct 2014 06:20:51 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1XjQMr-0002fr-QD for qemu-devel@nongnu.org; Wed, 29 Oct 2014 06:20:46 -0400 Received: from [59.151.112.132] (port=53918 helo=heian.cn.fujitsu.com) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XjQMr-0002f3-8g for qemu-devel@nongnu.org; Wed, 29 Oct 2014 06:20:41 -0400 Message-ID: <5450BF1F.4040801@cn.fujitsu.com> Date: Wed, 29 Oct 2014 18:19:11 +0800 From: Hongyang Yang MIME-Version: 1.0 References: <1411464235-5653-1-git-send-email-yanghy@cn.fujitsu.com> <54508EF5.8000908@cn.fujitsu.com> <20141029093457.GB2310@work-vm> In-Reply-To: <20141029093457.GB2310@work-vm> Content-Type: text/plain; charset="UTF-8"; format=flowed Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [RFC PATCH v2 00/23] COarse-grain LOck-stepping(COLO) Virtual Machines for Non-stop Service List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Dr. David Alan Gilbert" , Wen Congyang Cc: Gui Jianfeng , Jiang Yunhong , Dong Eddie , qemu-devel@nongnu.org, "Michael R. Hines" , Paolo Bonzini , Walid Nouri Hi Dave, =E5=9C=A8 10/29/2014 05:34 PM, Dr. David Alan Gilbert =E5=86=99=E9=81=93: > * Wen Congyang (wency@cn.fujitsu.com) wrote: > > > >> Hi all: >> >> I will start to implement disk replication. Before doing this, I think w= e should decide >> how to implement it. >> >> I have two ideas about it: >> 1. implement it in qemu >> Advantage: very easy, and don't take too much time >> Disadvantage: the virtio disk with vhost is not supported, because t= he disk I/O >> operations are not handled in qemu. >> >> 2. update drbd and make it support colo >> Advantage: we can use it for both KVM and XEN. >> Disadvantage: The implementation may be complex, and need too much t= ime to >> implement it.(I don't read the drbd's codes, and can't estimate= the cost) >> >> I think we can use 1 to implement it first. >> If you have some other idea, please let me know. > > For the COLO disk replication; are you talking here about 'local storage' > and treating it as 'internal state' or 'external state' (as described in = the > first half of 4.4 in the original COLO paper)? 'local storage' and 'internal state'. This is what later half of 4.4 said: 'COLO considers the state of local storage device as an internal s- tate of the guest, and snapshots the local storage state as part of the VM checkpoint, and plans to explore the other solution in future.' > > I know some groups would like to take advantage of facilities in the stor= age > layer to help; e.g. take snapshots/rollback etc - so I think it's best > to do (1) but make the interface clean so that other mechanisms could be = added. > Similarly I guess things like scsi-xcopy might help for some people - I'm > not saying implement these, just if possible make an interface where it c= ould > fit later. > > It's probably best to check with the QEMU storage guys that you can reuse > anything they have; there was a discussion a few weeks back where I cc'd > Fam, Stefan and Kevin in). > > Dave > > -- > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK > . > --=20 Thanks, Yang.