Cisco Support Community
Community Member

WAN redundancy/failover


I have remote site connecting to 2 hubs. I would like to set redundancy/failover between main connection T1 and a DSL backup.

How can I achieve this with a routing prrotocol 0SPF/EIGRP.


Hall of Fame Super Bronze

WAN redundancy/failover

Ideally, you want to run the same routing protocol on both WAN connections.

Depending on the routing protocol of choice, you can influence which link would be preferred by using 'ospf cost' or 'eigrp bandwidth or delay' metrics.

If your DSL is going over a public network (internet), you will need to configure a GRE tunnel and enable the routing protocol in the tunnel.



Community Member

WAN redundancy/failover

Thanks for your quick response.

I have DSL going over public network and getting negotiated IP address and using fastethernet interface.

for the GRE tunnels, should I anchor them using the FA or create a new loopback?

do you have any link that discuss such solution? I am not opting for any failover using static route or tracking. I like the failover to be dynamic.

Hall of Fame Super Bronze

WAN redundancy/failover

The GRE should form by using your FastE interfaces facing the DSL provider as the tunnel source.

The Tunnel interface will have a /30 subnet (private) and this subnet can be added to the routing protocol of choice.

If you want the tunnel to be used along with the T1 (load-balancing with CEF),

then the cost to/from sites must be the same.

With OSPF, you can manipulate this condition with the 'ip ospf cost' command.

If you want the tunnel to be used as failover/backup, then you can increase the cost on the tunnel interface

so the T1 will be primary.

You can do a internet search on 'ip ospf cost' and you will find many examples of its use and design guidance.



CreatePlease to create content