пт, 12 апр. 2019 г., 12:10 Ruben Kerkhof <ruben@rubenkerkhof.com>:
On Thu, Apr 11, 2019 at 10:43 PM Vasiliy Tolstov <v.tolstov@selfip.ru> wrote:

Hi Vasiliy,

>
> ср, 10 апр. 2019 г. в 19:46, Ruben Kerkhof <ruben@rubenkerkhof.com>:
> >
> > On Wed, Apr 10, 2019 at 1:44 PM Laine Stump <laine@redhat.com> wrote:
> > >
> > > On 4/9/19 11:35 AM, Ruben Kerkhof wrote:
> > > > On Tue, Apr 9, 2019 at 5:10 PM Michal Privoznik <mprivozn@redhat.com> wrote:
> > > >>
> > > >> On 4/9/19 4:38 PM, Ruben Kerkhof wrote:
> > > >>> Hi all,
> > > >>>
> > > >>> I have a hook script, /etc/libvirt/hooks/network, that doesn't seem to
> > > >>> be called when I attach an interface with type 'ethernet' with this
> > > >>> xml snippet:
> > > >>>
> > > >>> <interface type='ethernet'>
> > > >>>       <model type="virtio"/>
> > > >>>       <source>
> > > >>>           <ip address="10.100.0.1" prefix="24" peer="10.100.0.10"/>
> > > >>>       </source>
> > > >>> </interface>
> > > >>>
> > > >>> https://www.libvirt.org/hooks.html#intro says
> > > >>> "A network is started or stopped or an interface is plugged/unplugged
> > > >>> to/from the network (since 1.2.2)".
> > > >>>
> > > >>> While I don't have a network defined in xml, I'd expect this to work
> > > >>> just as well for 'ethernet' type interfaces. Am I wrong?
> > > >>
> > > >>
> > > >> Hotplugging an 'ethernet' type of interface doesn't really relate to any
> > > >> libvirt network. Hence libvirt doesn't call 'network' hook script. If
> > > >> you'd continue reading you'll see what is the 'network' hook fed with
> > > >> (on stdin): info on domain in question AND network where the event
> > > >> ocurred. But there is no network, is it?
> > > >
> > > > No not in the libvirt sense there isn't, you're right.
> > > >>
> > > >> But maybe you can work around this by waiting for
> > > >> DEVICE_ADDED/DEVICE_REMOVED events? What is it that you're trying to solve?
> > > >
> > > > I'd like to enable proxy_arp on the interface among other things.
> > > > I can easily do this from the same script that adds the interface
> > > > though, so I have a workaround, but a hook that triggers on all
> > > > interface events felt cleaner.
> > >
> > > Also keep in mind that the hook scripts aren't an officially supported
> > > part of the API, and are thus liable to change without warning. As an
> > > example, danpb has proposed changing the network hook:
> > >
> > >   https://www.redhat.com/archives/libvir-list/2019-March/msg01280.html
> > >
> > > Once this goes in, any network hook script that uses the plugged and
> > > unplugged hooks will no longer work; you would instead need to use the
> > > port-created and port-deleted hooks.
> >
> > That's good to know, thank you Laine.
> >
> > In that case perhaps adding this functionality to libvirt itself would
> > make more sense?
> >
> > Some background, I'm implementing
> > https://vincent.bernat.ch/en/blog/2018-l3-routing-hypervisor, so
> > announcing /32 host routes to our vms with BGP.
> > Our existing vms use /24 network addressing, and to keep that working,
> > I need to set proxy_arp on the interface.
> >
> I done something like this in some Russian Cloud hosting provider, but
> we use OSPF to redistribute routes in DC, and sync their via bgp
> between dc.

Thanks, it's good to know that more people do this and I'm on the right path :)

> I don't recommend hooks, write small netlink based program that
> listens events about interface up, check it names and set proxy arp
> for it.
> This is not blocking and easy to modify/test/write. Or set proxy_arp
> globally (but this is not recommended)

That's a good tip, thanks.
I do want to set proxy_arp before the interface is admin up and
passing traffic, is that possible with netlink?

Writing an udev rule might work as well now that I think of it.

But seeing that multiple people have implemented this in their own
way, wouldn't it make more sense to add support for this to libvirt?
Something like <interface type='ethernet' proxy_arp='on' />
>


This is not one option that can be setted, so i think this is not generic variant.
And yes,if this is only one option,udev is sufficiently.

> --
> Vasiliy Tolstov,
> e-mail: v.tolstov@selfip.ru

Kind regards,

Ruben