From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:33170) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UWiQF-0000Ba-4H for qemu-devel@nongnu.org; Mon, 29 Apr 2013 03:22:53 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UWiQC-00082A-AM for qemu-devel@nongnu.org; Mon, 29 Apr 2013 03:22:51 -0400 Received: from mx1.redhat.com ([209.132.183.28]:35306) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UWiQB-000821-Qc for qemu-devel@nongnu.org; Mon, 29 Apr 2013 03:22:48 -0400 Date: Mon, 29 Apr 2013 09:22:44 +0200 From: Stefan Hajnoczi Message-ID: <20130429072243.GB13488@stefanha-thinkpad.redhat.com> References: <1366734308-11724-1-git-send-email-stefanha@redhat.com> <517B6403.5020100@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <517B6403.5020100@linux.vnet.ibm.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH 0/3] block: block-backup live backup command List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Wenchao Xia Cc: Kevin Wolf , Fam Zheng , qemu-devel@nongnu.org, Ian Main , Paolo Bonzini , dietmar@proxmox.com On Sat, Apr 27, 2013 at 01:37:07PM +0800, Wenchao Xia wrote: > =E4=BA=8E 2013-4-24 0:25, Stefan Hajnoczi =E5=86=99=E9=81=93: > > This series adds a new QMP command, block-backup, which takes a point= -in-time > > snapshot of a block device. The snapshot is copied out to a target b= lock > > device. A simple example is: > >=20 > > block-backup device=3Dvirtio0 format=3Dqcow2 target=3Dbackup-20130= 401.qcow2 > >=20 > > The original block-backup blockjob was written by Dietmar Maurer > > . He is currently busy but I feel the feature i= s worth > > pushing into QEMU since there has been interest. This is my version = of his > > patch, plus the QMP command and qemu-iotests test case. > >=20 > > How is this different from block-stream and drive-mirror? > > --------------------------------------------------------- > > Both block-stream and drive-mirror do not provide immediate point-in-= time > > snapshots. Instead they copy data into a new file and then switch to= it. In > > other words, the point at which the "snapshot" is taken cannot be con= trolled > > directly. > >=20 > > block-backup intercepts guest writes and saves data into the target b= lock > > device before it is overwritten. The target block device can be a ra= w image > > file, backing files are not used to implement this feature. > >=20 > Besides, compared to snapshot-blkdev, it mainly brings better > performance by avoid merging later, however, other tool may > be needed to form an incremental backup, which may be not related to > this patch. > No objection to this patch, but perhaps a better way is using > internal snapshot by adding base/delta data export support. Yes, incremental backups would need to be added later. Stefan