Hi,
On Tue, Apr 08, 2025 at 12:17:45AM +0200, Tobias Fiebig wrote:
> This occurs, for example, if you have an RFC8950 session to a route-
> server at an IX and the route-server also sends you v4 routes with a
> _v4_ nexthop.
I would consider this to be a bug. If the route-server knows (by
means
Tobias Fiebig wrote on 07/04/2025 23:17:
This occurs, for example, if you have an RFC8950 session to a route-
server at an IX and the route-server also sends you v4 routes with a
_v4_ nexthop.
At the risk of raining on some parades, rfc8950 at internet exchanges is
experimental so far. The lon
Thanks for the suggestions! I posted a new version of the document based on
this and other suggestions from IETF 122:
https://datatracker.ietf.org/doc/draft-karstens-intarea-multicast-application-port/
Rather than provide specific guidance to the review team, the new version just
points out tha
Chairs,
This draft was discussed by pim during IETF 118 & 120 and intarea during IETF
121 & 122. There have also been some reviews on the mailing list. Would this be
an appropriate time to consider adoption by the intarea WG?
https://datatracker.ietf.org/doc/draft-karstens-intarea-multicast-app
Moin,
> I think the number of hosts and routers on the global Internet which
> have IPv6 enabled and IPv4 disabled can probably be counted on one
> person's fingers and toes; see:
>
>
> https://ungleich.ch/en-us/cms/blog/2019/02/05/list-of-ipv6-only-services/
> https://sites.ip-update.net/
N
Moin,
> You mean a router, right? If a host has no IPv4 addresses assigned,
> then I don't see how it can send an IPv4 packet.
Actually; *on an interface.
one way to use v4-w-v6 nh in practice is binding the v4 address to lo
(granted, on routers, but this also holds for packets they originate);
Moin,
> At the risk of raining on some parades, rfc8950 at internet exchanges
> is experimental so far. The long term future of this approach isn't
> clear, given that its intention is to relieve the impact of lack of
> availability of ipv4, which is readily available at modest cost, at
> least r
Moin,
On Tue, 2025-04-08 at 12:31 +0200, Gert Doering wrote:
> I would consider this to be a bug. If the route-server knows (by
> means of config) that my router wants to receive routes with v6
> nexthops only, because my router does not *have* a v4 address, it
> must not send me something know
>> Could you please describe the problem you're solving?
> A host that 'usually' only does v4-w-v6 nexthop and does _not_ have
> a(ny) v4 addresses getting routes with a v4 nexthop that are on a link
> on which the host also is (but only has a v6 address on).
You mean a router, right? If a host
Tobias Fiebig wrote on 08/04/2025 14:04:
...is kind of useful here, because it also gives an IX an easy way to
do proxy "ARP" for those v4 members, without actually sending ARP that
would be misunderstood by them.
Now I've seen "IX" and "proxy arp" in the same sentence.
As a proposal.
Can we
Moin,
> Now I've seen "IX" and "proxy arp" in the same sentence.
>
> As a proposal.
>
> Can we take the entire idea away and throw it into the nearest
> volcano?
To be fair to the idea: DE-CIX iirc already does, effectively, proxy
ARP and NDP for all members. ;-)
With best regards,
Tobias
--
11 matches
Mail list logo
