Hello Steve,
is the redistribution of BGP into EIGRP taking place on the same device or on the neighbor device or both are redistributing the same BGP prefix?
about the next-hop, if I understand correctly the IP 10.46.246.14 is that of a regular EIGRP neighbor?
Is correct to say that this output is taken by a debug ip eigrp events rather then a show command?
There are some non-sense messages in EIGRP debug messages like when it complains to receive its own EIGRP hellos on loopback interfaces!
I would focus on EIGRP topology table for the IP prefix in question does it show the correct expected ouput?
sh ip eigrp topology
I wonder if you are seeing the signs of a competition between two EIGRP nodes in injecting an external route for the involved prefix or the result of a route filter applied outbound on local node on interface receiving the unwanted update.
or if even a neighbor is sending on an interface an update that contains an external route generated by local node passed to a third node and then sent to this local node by the above EIGRP neighbor (because received on another interface)
if so local node receives its own update back on another interface and complains of having recevied an update that looks coming from a duplicated node.
may you check if the prefix is blocked by a distribution list out the interface receiving the wrong update?
if so the neighbor tries to send an update to the local node because it is no receiving the same prefix on the link.
EIGRP router-id is places in a field of external EIGRP routes and allows for a consistency checks before accepting an external route.
But in general EIGRP builds neighborships on per link basis and cannot recognize the same node as OSPF or ISIS could do.
Edit:
the Fas0/2 is in shutdown ? well this makes the whole thing too strange.
Hope to help
Giuseppe
