Good day,
I'm doing some lab preparation for my CCIE exam and I would just like to understand the concept of sham link better. I've attached a part of the lab scenario i've done recently (pardon for the low quality).
The scenario is that, I have a DMVPN connection between 3 sites and is running ospf with network type point to multipoint. As the network is still using the dmvpn path, the routing table is as shown below (all host IPs are reachable).
@R1
O 172.16.30.3/32 [110/1000] via 172.16.30.3, 00:01:50, Tunnel1
O 172.16.30.5/32 [110/1000] via 172.16.30.5, 00:11:29, Tunnel1
When the MPLS was then configured and was used as the primary path (using sham link) for the reachability of LANs between R1 & R3, the DMVPN host ip was already unreachable (.1 unreachable from R3, .3 unreachable from R1, .5 is reachable on all routers). As shown in the new routing table, the next hop then changes because of the sham link configuration.
@R1
O 172.16.30.3/32 [110/100] via 172.16.3.1, 00:01:25
O 172.16.30.5/32 [110/1000] via 172.16.30.5, 00:14:44, Tunnel1
where 172.16.3.1 is the destination sham link ip, advertised via mpbgp.
The ospf neighbor relationship on the DMVPN was not affected and LANs then traverse through the MPLS path.
I tried filtering the host route so the router will then use its Connected entry in the routing table to reach the DMVPN host IP and it worked as expected (can ping).
I would like to understand why it wasn't able to ping the other host IP using the sham link as the next hop. And if this is the case, what would i do to have full reachability? Should I just remove the host route entry?
Thank you very much.
