Joakim Tjernlund wrote:
-----Original Message-----
From: Richard Ogier [mailto:og...@earthlink.net]
Sent: den 8 januari 2009 21:40
To: joakim.tjernl...@transmode.se
Cc: ospf@ietf.org
Subject: Re: [OSPF] Unnumbered PtoP router LSA question.
Jocke,
Here is one more piece of evidence (to add to your already convincing
evidence below), which makes it clear that stub links are required
for point-to-point links only when interface addresses are assigned.
The following is from RFC 2328, page 15.
"Interfaces to point-to-point networks need
not be assigned IP addresses. When interface addresses are
assigned, they are modelled as stub links, with each router
advertising a stub connection to the other router's interface
address. Optionally, an IP subnet can be assigned to the point-
to-point network. In this case, both routers advertise a stub
link to the IP subnet, instead of advertising each others' IP
interface addresses."
There is no longer any doubt in my mind about this. The main
clarification is that Footnote 2 MUST be implemented.
One can always add stub links that are not required.
Richard
Thanks
I do think however that Dave has a point. There are probably impl.
out there that does Option 1, but does not include a host route
to itself. Such routers will probably have a hard time working
with an router not doing Option 1.
I suspect that footnote 2 isn't enough, one must announce an host
route as soon as one have 1 or more unnumbered links.
Jocke
If your suspicion is correct that Footnote 2 isn't enough, then that
would be an error in RFC 2328, since as we seem to agree, the RFC
does not require doing Option 1 for unnumbered links.
However, I am not sure your suspicion is correct. Do you have an
example that shows Footnote 2 is not enough to ensure each router is
reachable? I agree that if one wants to ensure *optimal* routes to
each router, then a router must announce a host route when it has 1
or more unnumbered links. (I can give an example to show this.)
But there is no requirement of optimal routes for management purposes.
Richard
_______________________________________________
OSPF mailing list
OSPF@ietf.org
https://www.ietf.org/mailman/listinfo/ospf
