Hi Matthias,
On Tue, 5 Nov 2013, Matthias Waehlisch wrote:
actually, I'm confused about the statemet on slide 4: "Different
validation states only happens when there is a less or more specific
route announced."
For example, ROA {1.2.0.0/16-24, AS 10}. BGP Update 1.2.0.0/16 with
origin AS 10 is valid, but the same prefix with origin AS 20 is invalid.
Can you clarify?
Yes, I'm trying to figure out in what situation a valid route has to be
prefered over an unknown route, invalid routes are allready dropped by the
validation routing-policy.
Say route A is announced (and isn't invalid).
1. This route will become Unknown when route A is not covered by a ROA. This
Unknown route can either have or not have more-specific routes being
announced. If no more specific routes are announced this is the best route
(no preferring necessary).
2. If there are more specific routes for route A and these are covered by a
ROA they will be Valid (invalids are already dropped). But since they are
more specific to route A prefering them above the Unknown (less specific)
route A makes no sense.
3. If there are more specific routes for route A and these aren't covered by
a ROA they will be Unknown, goto 1.
Knowing this, I can't think of a reason to prefer valid routes over
unknowns and that makes defining a routing policy (as shown in slide 8) a
whole lot easier.
Jac
Thanks
matthias
--
Jac Kloots
Network Services
SURFnet bv
_______________________________________________
sidr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sidr
