[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: A question about VM configuration


  • To: xen-users@xxxxxxxxxxxxxxxxxxxx
  • From: Paul Leiber <paul@xxxxxxxxxxxxxxxx>
  • Date: Fri, 23 May 2025 18:23:18 +0200
  • Arc-authentication-results: i=1; strato.com; arc=none; dkim=none
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; t=1748017410; s=strato-dkim-0002; d=strato.com; h=In-Reply-To:From:References:To:Subject:Date:Message-ID:Cc:Date:From: Subject:Sender; bh=yuUMWhmcT+MMcH6jw80dLIapZ89+OL0HxhTP5ctM0fU=; b=i3E2AwJOSAA6pUD2Py02ln8c/Js9gUP05IwaNiYuOpHdA1FE+KMhxZEX5Q2ZuhyuwM aN/MoxPv8pV/WwpNCzi+rXKB/sg41TzoevMPtz6PFH+IVlEtID+NQVZ+qm0XO3xTZzov 8rLieADdTNrQ65klBhmQrrY4Qd+Uwc2OTCrmMIiLt2qmLLCDIXHldec5HeLbgD97uHTV khpIJ+klNkjXv4U6+nMiSJ284bC9hr5aB4DVwdfRsft7uN/HxxnqhD5FhfrBtbJ7n+2R ph9RQRblZtlp+LyRP56JPWL4R6prYRx3hIUddfaNz1dH0ejcaZflJr/fr0bapbarrX03 vnUw==
  • Arc-seal: i=1; a=rsa-sha256; t=1748017410; cv=none; d=strato.com; s=strato-dkim-0002; b=YKsO9fnohuYSUJeqQf1FV6VFqX6Z8+7NlI80kTXbgNPAI/gGM00iZT7bpEt5a2MEf/ uiZmexpraz+sVGqzp3RQSFuXux6iZ7JCFhcE6CVV2BrrqaGXrXQMv7j3AzpqbmJ4aO08 ZAaSXIt9EG25AfASU4CtT74TxLEPr+ElPh99aMRjT+v4JUOHh+wMSs8aY/Sxs+2PxGIu WlXZGcguO5J0sIixg1lOGpIiuNB0GdhEVB7wF8lgXQaTf54yYN7EmNsYx66DEgUA62KS diI7OQvvO05aiIZurBMaTZgAh7yT6B4K9wWhoZ12vYt42+uX5NbWjiy9XAfbEmHUBsdg XF3A==
  • Delivery-date: Fri, 23 May 2025 16:24:31 +0000
  • List-id: Xen user discussion <xen-users.lists.xenproject.org>

On 23.05.25 8:15 AM, Mario V. Guenzi wrote:

Hello Attila,

T.Y.

The problem isn't transferring or cloning the VMs — I can do that without any issues using |dd| or Clonezilla. The real problem, as Paul mentioned yesterday, is that there's a bug in OVMF which has existed since 2013 and still hasn't been fixed.

Mario, I'm not sure we are talking about the same bug. The bug in the Debian ovmf package is caused by architectural changes in ovmf not reflected in the Debian package, namely the creation a xen-specific ovmf version [1] which doesn't exist in Debian ovmf package 2022.11-6. (The change in the code happened in 2019, so I am not sure which bug you are referring to which is supposed to exist since 2013.) Lacking this ovmf version, uefi boot of xen DomUs fails. Other distributions are including this xen-specific ovmf version, Debian doesn't (yet?). Installing the previous Debian ovmf package is sufficient to circumvent this problem. building the right ovmf version and using it fixes this problem (more details can be found in the bug report and replies). As Devuan is based on Debian, I thought it might be probable that this bug has been taken over by Devuan.

To make that clear: I am in no position to know precisely what the cause of your issue is, I am just making assumptions. The actual root cause might be something else.

So, when I start the VM with the |firmware = 'ovmf'| or |uefi| option (which are essentially the same), it crashes and reboots after a few seconds.

Now I need to find a way to create a Windows 10 ELTS machine that can boot without UEFI firmware, and then, once it's up and running, try cloning only the partition I need from the old VM.

The issue is that even using the configuration parameters Joos gave me yesterday, the machine doesn’t work — it doesn’t even boot from the CD-ROM.

You could consider trying one of the workaround/fixes mentioned above  (using an older ovmf version or building ovmf yourself). This might be less effort than the path you describe, but of course YMMV. (Switching to a previous ovmf version is one shell command, at least in Debian. If this change is effective, this would also rule out other causes at the same time.)

Paul

[1] https://lore.kernel.org/all/20190813113119.14804-1-anthony.perard@xxxxxxxxxx/#t





 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.