You have probably seen this route in the routing table by now and wondered, what is it doing there. Then you do some research and read that the route has been added to avoid routing loops. But let’s actually see why and when would such routing loop occur!
I will use the following topology:
- R1 is the gateway for three departments, X, Y & Z
- R1 is configured with a static default route, pointing to router R2
- R1 and R2 exchange routing via EIGRP
The configuration here is very simple – so if we look at the routing table on router R2, we’ll see the following routes installed:
In order to optimise the routing and to minimise the routing table, we will configure R1 to advertise a summary instead; this will include all the EIGRP routes above … So once the configuration is done, the routing table on R2 looks like this:
Ok … so now, let’s run a traceroute from R2, towards a network that doesn’t exist, yet covered by the summary route:
When R1 gets the packet, it won’t find a route in its routing table – so it will forward the packet back to R2 as per its default route! And this is where the packet will start looping between R1 and R2 until TTL=0.
EIGRP will by default install a route pointing to Null0 interface with a Administrative Distance of 5 whenever a summary route is advertised out. In this case, I have disabled that behaviour by setting the Admin Distance to 255 (infinite).
This feature is also available with OSPF and BGP, but not with RIP – however, should you need similar functionality, nothing would stop you from adding a static route pointing to the Null0 interface.