Route Tagging to Stop Routing Loops Causing by Multi-point Redistribution

Route tagging is a mechanism which is designed to help to mitigate routing loops which can be caused due to multi point redistribution. As a best practice, if you are configuring a network where there are multiple points of redistribution, you should implement Tagging. Route Tagging can also be used to identify routes which then later be used in a route-map for further traffic engineering.

Looping case with multi point redistribution.

All the ip addresses of the interfaces are like 192.168.XY.[X|Y]/24 where X and Y are router numbers. As an example, R3-R4 link IP address of R3 (e0/1) is 192.168.34.3/24

When the R1 advertises it's loopback subnet of 1.1.1.0/24 to RIP it will be learned by R4 and R6.
After the route is redistributed from RIP to OSPF, R6 will learn it from OSPF too and if the OSPF routes are redistributed to RIP at R6 with a lower metric, (let's say metric of 1) it will be advertised to R3 via RIP. So now R3 see the 1.1.1.0/24 is learning from R6 with a low metric route than R2 and R3 select the R6 as the best path to reach 1.1.1.0/24

Now here is the loop. If R5 takes a traceroute to 1.1.1.1 it will be like the following..

This happens because the routes are redistributed back to the originated routing protocol and of course in this case with a lower metric.

In order to prevent these kind of routing problems from happening, we will use tagging with a route-map like the following and combine it with the redistribution command.

R(config)#route-map TAG deny 10
R(config-route-map)#match tag 1
R(config)#route-map TAG permit 20
R(config-route-map)#set tag 1

When used with a redistribute command, above route-map says not to redistribute tagged traffic with tag-id 1 & tag with tag-id 1 and redistribute all the traffic which are not tagged with tag-id 1.

You will have to do this on both routing protocols on both routers..