We are currently having some issues with running ospf over two tunnel interfaces. Both of these tunnels are originating from the same interface, and thus public IP, but both tunnels have different private ips on both endpoints. It is because of this that we set up the tunnels with the ospf network type point to point, but we are having some issues. It appears that the router with both tunnels originating from the same interface is not seeing itself in hellos from the other two routers, and the other two routers are continually retransmitting DBD packets to no avail. I have already checked the basic things, and all of the parameters for ospf neighborship are matching. I have been doing some debugging and here is something interesting:
*Mar 22 21:04:38.828: OSPF: Rcv DBD from <hub router> on Tunnel3 seq 0x1F57 opt 0x52 flag 0x2 len 1392 mtu 1476 state EXSTART
*Mar 22 20:20:15.309: OSPF: Rcv DBD from <endpoint router> on Tunnel3 seq 0xE4E opt 0x52 flag 0x7 len 32 mtu 1476 state EXSTART
After some reading online, it appears that the Flag's being different signifies that the routers are assuming the network type is different. According to the article that I read, the router with the two tunnels originating from the same interface wants this network to be set up as point to multi-point, but this doesn't seem to make sense as each tunnel should be its own point to point interface. I am curious if anyone has any insight, and or experience with this issue.
I can provide additional information if needed, and thanks in advance for any help.
Andrew
