Hi Sam, | In section 2, "covering" and "matching" are used without definition. | What do YOU mean by covering in this context? In the document, please | define or cite a definition. | | In section 5, repeat the informative reference to | [I-D.ymbk-rpki-rtr-protocol] as a suggested protocol.
Ack on both of these. I am updating the draft to issue a new version as we speak and will take care of these. | How does the route preference (preferring "valid" over "not found" | over "invalid") interact with the preference for the longest prefix? | If I have an invalid /16, does it still take precedence over a valid | /8? (That may be another way of getting at the problems discussed in | the thread about this doc in November 2008.) Yes. And my answer is going to be the same ;-) The /16 and /8 are different routes/paths and the validity state is determined separately. If all the paths for the /16 route on a router are invalid, the route itself is not going to be present in forwarding (barring knobs of course). In that case, yes, the /8 route will be effective for all traffic that would have otherwise matched the /16. - Pradosh _______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
