OSPFv3 doesn't announce loopback IPv6 network

[Bernhard Ehlers]

On 11.12.22 21:50 Bernd Naumann via Bird-users wrote:
> On 11.12.22 18:50, Ondrej Zajicek via Bird-users wrote:
>> On Sun, Dec 11, 2022 at 06:30:46PM +0100, Bernhard Ehlers via 
>> Bird-users wrote:
>>> I want to use OSPF with BIRD v2.0.10 on Linux.
>>>
>>> That works well, except that the IPv6 network of the lo interface isn't
>>> announced on OSPFv3. The IPv6 networks of the ethernet interfaces are
>>> announced as well as the IPv4 networks of all interfaces (ethernet 
>>> and lo).
>>> The only missing is the IPv6 network of lo.
>>
>> Hi
>>
>> Due to some implementation details, OSPFv3 ignores interfaces that do not
>> have link-local addresses (even if they are just stub). Just use:
>>
>> ip address add fe80::1/128 dev lo
>>
> 
> May I ask why not using the direct protocol to get the loopback addresses?

I tried that as well and it worked.

But I was concerned by this part of the documentation of the direct 
protocol:

> The question is whether it is a good idea to have such device routes
> in BIRD routing table. OS kernel usually handles device routes for
> directly connected networks by itself so we don't need (and don't
> want) to export these routes to the kernel protocol. OSPF protocol
> creates device routes for its interfaces itself and BGP protocol is
> usually used for exporting aggregate routes.

So for OSPF the direct protocol should not be necessary and indeed it 
worked without it for the ethernet interfaces (both for IPv4 and IPv6) 
and for the lo interface on IPv4. Therefore my question.

I think it is just personal choice, if one prefers adding the link-local 
address or using the direct protocol.