Hi Lucas,
It is working perfectly, until I'm restarting my network (using ifreload,
ifup, systemctl network restart, ...) - all my VM come unreachable...
To
make it work again, I also need to restart libvirtd
Is it still expected to have this behavior ?
Yes, I think it's expected, and the scenario is similar with
https://bugzilla.redhat.com/show_bug.cgi?id=1730084#c36
What would be a better way to configure the network ?
Refer to the
https://libvirt.org/formatdomain.html#bridge-to-lan, it is
recommended to use Linux bridge instead of ovs bridge.
Thanks.
BR,
Yalan
On Mon, Dec 30, 2024 at 9:42 PM linux--- via Users <users(a)lists.libvirt.org>
wrote:
> Hi,
>
> First, sorry if the topic has already been discussed recently (the only
> thread I found related to my problem was created in 2010..)
>
> I have a hypervisor with KVM and LXC installed on a Debian 12, and I'm
> using libvirt to create my VM. All my networks are defined in my
> /etc/network/interfaces.d/* (I'm using openvswitch with specific options,
> such as port mirroring/patch/...) , and I'm configuring the network on my
> VM XML definition like this :
>
> <interface type='bridge'>
> <mac address='52:54:00:ab:c3:d3'/>
> <source bridge='prod'/>
> <vlan>
> <tag id='55'/>
> </vlan>
> <virtualport type='openvswitch'>
> <parameters
interfaceid='331d973c-0c5b-4d3c-b2ad-590f908e680d'/>
> </virtualport>
> <target dev='vnet180'/>
> <model type='virtio'/>
> <mtu size='9216'/>
> <alias name='net1'/>
> <address type='pci' domain='0x0000' bus='0x00'
slot='0x05'
> function='0x0'/>
> </interface>
>
> It is working perfectly, until I'm restarting my network (using ifreload,
ifup, systemctl network restart, ...) - all my VM come unreachable...
To
make it work again, I also need to restart libvirtd
>
> Is it still expected to have this behavior ? What would be a better way to
> configure the network ?
>
> Thanks in advance
>
> Lucas
>
>