Pls find the output from the "show ip route" during normal operation.

and config also

How have you tested this failover. If the connection is failing between the two microlan devices then the line protocol on the interface attaching to your routers may not be going down. If this is the case the router will think that the primary route is still active and will continue to send traffic over that link into a blackhole.

One possible solution here is to use a routing protocol over both connections, such as eigrp. This would enable the router to learn about the intermediate link failure and adjust its routing table accordingly. The static route with higher AD option will only work if the interfaces on both sides of the link are aware of the intermediate link failure.

You can use conditional based policy routing to accomplish this as well. It uses a ping responder from the IP SLA software to test availability of remote ip address. If the ip becomes unavailable (possibly from a service interuption) conditional based policy routing will change the next hop ip address to the backup link next hop. You can set the thresholds for failover time and such.

As noted before this is essentially tracking an object (in this case a rtr ping responder availability) and using that the test the validity of the next hop. This is very common in scenarios where customers cannot run routing protocols over the internet to their service provider but require next hop redundancy to the internet.

Hope that helps.

Mike Louis

Dear Mike ,

Im little bit confused , I saw the config in below hyper link, hope it will work but IP policy should be implemented on Vlan1(management Vlan) ?Pls suggest

http://www.cisco.com/en/US/tech/tk364/technologies_configuration_example09186a0080211f5c.shtml,

thanks

Pramod Prasad

is there any way that you can configuring ip routing protocol, like eigrp across both of these links? That would be a much easier approach.

Not sure what you mean by

"ip policy should be implemented on Vlan1"

can you clarify?

The policy routing used here is only for changing the next hop in the routing table. Just like a routing protocol on both ends would do. You are essentially saying

check availability of router 1 (primary remote device) using a ping check

if remote device responds use current default route and next hop

if remote device fails to respond then use this alternate default route and next hop

this is automating what an admin would have to do if the primary link failed in order to restore packet forwarding across the backup link. This configuration will do it for you.

i would really stress using a routing protocol if possible. The configuration is very simple. We could even use a tunnel configuration if service provider is a problem. Although we would have to adjust the mtu size on the tunnel but it would work and we could run routing protocols over the tunnel . Just another thought.

Mike

Hi Mike,

I got the point tht using the routing protocol is best option , but i have a constraint tht im not able do so i.e looking 4 the option tht will provide backup option using static routes which will be automating the backup process.

U r advicing the tunnel option but this option is feasible in Switch without using routing protocol? Is this option automate the backup process.

Pls suggesst.

Pramod Prasad

What ios version are you running on the wan routers?

At both ends there is 4507 catalyst switches , ONE LINK IS micro LAN first come to microlan bridge then extended to our catlyst , another link is by BG -40 first terminated at 3500 switch then extended to our catlyst.

both terminated on fastethernet of catalyst

.