From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [140.186.70.92] (port=33149 helo=eggs.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1PTWy2-0001wV-P9 for qemu-devel@nongnu.org; Fri, 17 Dec 2010 04:51:15 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1PTWy1-0008Sr-GZ for qemu-devel@nongnu.org; Fri, 17 Dec 2010 04:51:14 -0500 Received: from mx1.redhat.com ([209.132.183.28]:25268) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1PTWy1-0008Sm-8d for qemu-devel@nongnu.org; Fri, 17 Dec 2010 04:51:13 -0500 Message-ID: <4D0B32D3.5070507@redhat.com> Date: Fri, 17 Dec 2010 10:52:19 +0100 From: Kevin Wolf MIME-Version: 1.0 Subject: Re: [Qemu-devel] -snapshot References: <4D0A5770.40308@mail.berlios.de> In-Reply-To: <4D0A5770.40308@mail.berlios.de> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Stefan Weil Cc: Stefan Hajnoczi , qemu-devel@nongnu.org, Amador Pahim Am 16.12.2010 19:16, schrieb Stefan Weil: > Am 16.12.2010 18:45, schrieb Stefan Hajnoczi: >> On Thu, Dec 16, 2010 at 4:34 PM, Amador Pahim wrote: >>> Thank you for your answer. Just one more question: If, while my >>> "snapshot" vms are running, the main disk is modified by a non >>> "snapshot" vm? For example, installing some extra software.. this can >>> freeze vms or something? >> >> Correct, it is not safe to modify the base image while there is >> another disk image backed off it. >> >> The reason for this is that the image only needs to store the changes >> that were made on top of the base image. For anything which hasn't >> been modified it will go back to the base image and read data from >> there. >> >> If you modify the base image, then the filesystem in the base image is >> not longer what your image file was created from and you have an >> inconsistent view of the disk. It leads to odd behavior and is >> unsafe. >> >> Stefan > > There are useful scenarios where using the same disk > simultaneously from a snapshot vm and a real system > works. > > If you have a hard disk with a dual boot configuration, > it is sometimes useful to boot one configuration with > the real system, then start qemu and boot the second > configuration. > > Even booting the same configuration twice > (once with the real machine, once with qemu snapshot) > is sometimes useful and works to a limited degree. > It is a simple way to try new bootloader configurations > or other boot setups. Right, though this doesn't contradict what Stefan said. It only works because in fact you don't modify the parts that your guest reads. Kevin