I'm doing the provision from foreman and not ovirt. So when you provision
from Foreman you cannot do a run once as it is calling the api side of
ovirt.
There is also the fact that when this is done on just libvirt/kvm it works
correctly with the boot order that is PXE first then HD second after the
system has been provisioned.
On Apr 9, 2015 16:38, "Michal Skrivanek" <michal.skrivanek(a)redhat.com>
wrote:
On Apr 9, 2015, at 08:38 , Brandon Merjil <bmerjil(a)ken-ohki.org> wrote:
> Hello,
>
> I am testing provisioning from foreman 1.8.0 to ovirt 3.5 with both on
CentOS 7.1503 and have run into an issue after the system has been deployed
to ovirt.
> When the guest system reboots after the kickstart completes it loads the
PXE menu from a the same tftp server which tries to start localboot 0. It
then tries to boot based on the order that ovirt had set for the guest,
which is PXE first and HD second.
> The message I'm seeing is "Booting from local disk...
> No more network devices
> No bootable device"
>
> If I change the guest boot options to only the virtual HD the guest
boots up without an issue.
>
> I have also done a provision to a plain libvirt/kvm server and had no
issue after the reboot. Same foreman server and tftp server with PXE menu
localboot 0 option were used.
> Any ideas what I might be running into here, and any additional
information needed.
The boot order is fixed in VM configuration. For initial config/install it
would be recommended to use Run Once with changed boot order
Thanks,
michal
> Thanks.
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
>
http://lists.ovirt.org/mailman/listinfo/users