nanog mailing list archives

Re: RFC 5549 - IPv4 Routes with IPv6 next-hop - Does it really exists?


From: Saku Ytti <saku () ytti fi>
Date: Wed, 29 Jul 2020 14:40:43 +0300

Hey,

On Wed, 29 Jul 2020 at 14:26, Alejandro Acosta
<alejandroacostaalamo () gmail com> wrote:

https://blog.acostasite.com/2013/02/publicar-prefijos-ipv4-sobre-una-sesion.html
https://blog.acostasite.com/2013/02/publicando-prefijos-ipv6-sobre-sesiones.html

I did not like, difficult troubleshooting in case something goes wrong (however I can understand it's a nice feature 
to have and in might be useful in some scenarios).

Your experiment predates extended nexthop encoding, but otherwise it
is indeed the very same thing. Just less operational overhead now.

Of course everyone has done 6PE and 6VPE longest time, because
obviously you can fit IPv4 next-hop in IPv6 coding, so nothing was
needed. This extended nexthop encoding only exists to fix the problem
that wire-format didn't support signalling IPv6 next-hop for IPv4
NLRI.

-- 
  ++ytti


Current thread: