Showing results for 
Search instead for 
Did you mean: 

Cisco 1941 Ethernet Interface Question


Here's my problem.  I'm going to be using Cisco 1941 routers at a bunch of remote sites.  All of these sites have 2 comm paths out.  Some of them have 2 IP/VHF radios and some have 1 IP/VHF radio and a copper link using Patton ethernet extenders.  From the VHF radios the data hit our MPLS network back to our HQ and the sites with copper go directly back to our HQ.  Everything ends up at a Cisco 4948 switch. 

The problem I'm having is that I want the routers at the remote site to use one ethernet port (G0/0) as the primary and the other (G0/1) as the backup interface.  I've tried the backup interface command but the problem is that depending on where an outage occurs the ethernet link to either the radio or Patton stays up so it never switches over. 

We're using OSPF as our routing protocol and I'm sure there's something that can be done with it but I'm not sure what. 

Does anyone have any recommendations?

Thanks in advance!


1 Accepted Solution

Accepted Solutions

Ok I found part of our problem with this issue.  The radio interface was being sent to a layer 2 switchport.  I switched it to a layer 3 instead and that fixed it.  I changed the OSPF priorities and tested it and it failed over as it should have and then returned back to it's primary path when it was restored.  That's exactly what I was trying to accomplish. 



View solution in original post

4 Replies 4

Richard Burts
Hall of Fame Guru Hall of Fame Guru
Hall of Fame Guru


There are several options that you could consider. I would suggest that the most simple (and therefore desirable) would be use use features in OSPF. You should be able to manually configure OSPF cost on the interfaces to change from the default values. Put a higher cost on the backup interface, a lower cost on the primary interface. Then OSPF should prefer to use the lower cost path. And OSPF will be able to detect when it no longer sees the neighbor if the primary path stops working (even if the Ethernet interface stays up the OSPF neighbor will become lost) and failover to the backup path should be automatic.

You could also consider using features such as interface tracking in IP SLA. This is frequently used in situations where there are static routes using the Ethernet interfaces. The fact that the Ethernet interface may still be up but there is no communication is a problem for the static route and IP SLA can recognize this. But it is more complex than using OSPF, so I suggest that you focus on the OSPF solution.






Thanks for the reply.  We have tried this with the cost and it doesn't seem like it's working either.  Do you need to change the cost on both ends of the link?  Do static routes affect this?  Somewhere along the line a previous network engineer here has done something to our network and we have to have static routes on our routers or they will not talk.  I've tried to make the static route point to the interface I want as the primary and that doesn't work so well either.