> OSPFv2, as specified in RFC 2328, states that a multi-access network's subnet > is specified by the DR's Network-LSA LSID/mask and installed during the SPF > graph traversal. Hence, without some form of signaling, there is no way to > prevent other routers in the OSPF area from installing the subnet route > (whether or not it is really needed for any purpose in the OSPF network > deployment).
So, in reality, this isn't about the way SPF is run, but the assumption in RFC2328 that the LSID/subnet mask from the type 2 is the "correct" subnet to put in the table when building the tree... Hmm... Then it seems like the magic number is unavoidable in this case. :-) Russ
signature.asc
Description: OpenPGP digital signature
_______________________________________________ OSPF mailing list [email protected] https://www.ietf.org/mailman/listinfo/ospf
