[ovirt-users] Using upstream QEMU

Michal Skrivanek michal.skrivanek at redhat.com
Fri Feb 2 14:10:14 UTC 2018



> On 1 Feb 2018, at 13:43, Harry Mallon <Harry.Mallon at codex.online> wrote:
> 
> Apologies for the formatting in the following message, I can't get Office for Mac to play ball...
> 
> 
> Harry Mallon
> CODEX | Senior Software Engineer
> 60 Poland Street | London | England | W1F 7NT
> E harry.mallon at codex.online <mailto:harry.mallon at codex.online> | T +44 203 7000 989
> On 01/02/2018, 09:28, "Michal Skrivanek" <michal.skrivanek at redhat.com <mailto:michal.skrivanek at redhat.com>> wrote:
>> On 31 Jan 2018, at 16:53, Yedidyah Bar David <didi at redhat.com <mailto:didi at redhat.com>> wrote:
>> 
>> On Wed, Jan 31, 2018 at 5:43 PM, Harry Mallon <Harry.Mallon at codex.online <mailto:Harry.Mallon at codex.online>> wrote:
>>> Hello all,
>>> 
>>> Has anyone used oVirt with non-oVirt provided QEMU versions?
>>> I need a feature provided by upstream QEMU, but it is disabled in the oVirt/CentOS7 QEMU RPM.
> 
>    just curious - which one?
>    Typically the reason for disabling it is that it’s not really stable
> 
> I am trying to run OSX guests on a host (Apple hardware). The "applesmc" device is part of that puzzle and is disabled in the Red Hat QEMU.
> 
>>> 
>>> I have two possible methods to avoid the issue:
>>> 1. Fedora has a more recent QEMU which is closer to 'stock'. I see that oVirt 4.2 has no Fedora support,
>> 
>> Indeed, mostly
>> 
>>> but is it possible to install the host onto a Fedora machine?
>> 
>> Didn't try this recently, but it might require not-too-much work with
>> fc25 or so.
>> IIRC fc27 is python3-only, and this will require more work (which is
>> ongoing, but
>> don't hold your breath).
>> 
>>> I am trying to use the master branch rpms as recommended in the "No Fedora Support" note with no luck currently.
>> 
>> Another option is to try to rebuild the fedora srpm for CentOS 7.
>> 
>>> 2. Is it safe/sensible to use oVirt with a CentOS7 host running an upstream QEMU version?
>> 
>> No idea. If it's only for development/testing, I'd say give it a try.
> 
>    it’s as sensible as running any bleeding edge stuff. It should work if you manage to resolve qemu deps
> 
>    you could also rebuild qemu-kvm-ev without the patch which blacklists the feature you want to see.
> 
> I was able to patch and rebuild qemu-kvm-ev, but I think I have hit more problems using the patched firmware from here: http://www.contrib.andrew.cmu.edu/~somlo/OSXKVM/ <http://www.contrib.andrew.cmu.edu/~somlo/OSXKVM/>. Hopefully that gets me a little closer though. Getting these mac VMs to work has so far been a huge pain.

I did try that couple years ago, it did work eventually, but I wasn’t able to get accelerated graphics and without that it was kind of useless

> 
>    Thanks,
>    michal
>> 
>>> 
>>> Thanks,
>>> Harry
>>> 
>>> 
>>> Harry Mallon
>>> CODEX | Senior Software Engineer
>>> 60 Poland Street | London | England | W1F 7NT
>>> E harry.mallon at codex.online | T +44 203 7000 989
>>> _______________________________________________
>>> Users mailing list
>>> Users at ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>> 
>> 
>> 
>> --
>> Didi
>> _______________________________________________
>> Users mailing list
>> Users at ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
>> 
>> 
> 
> 
> Harry

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20180202/a38f822a/attachment.html>


More information about the Users mailing list