Yes this is definitely problem of libvirt! However I think this will
take really long time to get fixed in libvirt. The patch is not a big
deal IMO and can be removed when libvirt fixes this. Also it does not
change the prior behavior.
-Jan
Fri, Apr 17, 2015 at 08:08:36AM CEST, olichtne(a)redhat.com wrote:
Isn't this a problem on libvirts side?
I'm fine with applying this patch (maybe mark it as a workaround) but
we should try to get this fixed in libvirt as well. What do you think?
-Ondrej
On Thu, Apr 16, 2015 at 06:34:20PM +0200, Jan Tluka wrote:
> The proper driver name for virtio devices is virtio_net. However libvirt
> does not accept it as value for model option therefore we have to treat
> it internally.
>
> Signed-off-by: Jan Tluka <jtluka(a)redhat.com>
> ---
> lnst/Controller/VirtUtils.py | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/lnst/Controller/VirtUtils.py b/lnst/Controller/VirtUtils.py
> index 25a6f39..19deee3 100644
> --- a/lnst/Controller/VirtUtils.py
> +++ b/lnst/Controller/VirtUtils.py
> @@ -90,6 +90,9 @@ class VirtDomainCtl:
>
> def attach_interface(self, hw_addr, net_name, driver="virtio"):
> try:
> + if driver == "virtio_net":
> + driver = "virtio"
> +
> device_xml = self._net_device_template.format(hw_addr,
> net_name,
> driver)
> --
> 1.9.3
>
> _______________________________________________
> LNST-developers mailing list
> LNST-developers(a)lists.fedorahosted.org
>
https://lists.fedorahosted.org/mailman/listinfo/lnst-developers