Cisco IOS Tunnel problem (or is it?)

[d.tabligan]

Hi,

I've been stumped for a few days already. Here's the scenario.

Router A is directly connected to Router B via E1 v.35 interfaces (link

C).

Router A is connected to Internet at point A v.35

Router B is also connected to Internet at point B v.35

Router A has equipments connected to Fast Ethernet port which includes

Lucent TNT, 2 Win2000 servers.

A tunnel is created between Routers A and B via their respective

internet connection points. Each is using it's own serial interface ip

(upstream internet provider) as source with the other router's serial

interface ip as the destination for the tunnel interfaces. Eigrp is

running between both routers on both serial interfaces as is iBGP to

aggregate the routes. When link C is up traffic is ok and eigrp prefers

link C due to its lower cost. All equipments connected to Router A via

its fast Ethernet are reachable. When link C goes down, the tunnel

interface should reroute via the Tunnel interface which it does fine, as

eigrp should do. Problem is, I cannot ping the Win2000 servers on the

Fast Ethernet segment in Router A from Router B, even though I can ping

the Lucent on the same segment. When link C goes back up, the problem

goes away.

What gives here?

Thanks in advance for your ideas!

Drix

[a-vazquez]

Does Router A ever receive the packets? Also, Is the problem only for Ping or accross the board? If both the devices(TNT and Win2k) are on the same physical segment and belong to the same subnet, and if the router A gets the packets but for select services only, the problem is most likely with access-lists. While pinging over the internet, the serial interfaces address is carried and might have been denied previously for this particular pair.