> 1. Text should be added to strongly recommend that when a route that is > about to expire is having an update of the expiration advertised that > receiving peers should treat the reception of an update with no other > changes to the reachability than the expiration time and signatures as a > refresh of the existing route. Implementations supporting temporal time > breaking in path selection should *not* treat the refresh as a new update.
I would think the problem here is that the BGP implementation must: 1. Know not to know run bestpath on such a path change. 2. Know to send a new update, although the bestpath hasn't changed. It would be useful to indicate the portions of the BGP RFCs that must be changed in order to support this. :-) Russ
signature.asc
Description: OpenPGP digital signature
_______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
