Routing metric
If a router learns about multiple different paths to the same network from the same routing protocol, a measure called metric is used to decide which route will be placed in the routing table. Just like with the administrative distance, the lower number represents the better route.
Each routing protocol has its own way to calculate the metric; Routing Information Protocol (RIP) uses hop counts, OPSF uses a parameter called cost, EIGRP uses bandwidth and delay to compute the metrics, etc. Note that metrics of different routing protocols can not be directly compared – an EIGRP route might have a metric of 4.042.334, while a RIP route can have a metric of 3.
If two routes have the same AD as well as the same metrics, the routing protocol will load-balance to the remote network, meaning that data will be sent down each link.
This is the another method used by the distance vector routing protocols to prevent routing loops called as route poisoning. Ehen a router defects that one of its directly connected routes has failed, it will advertise a failed route with an infinite metric (“poisoning the route”). Routers who receive the routing update will consider the route as failed and remove it from their routing tables.
Each routing protocol has its own definition of an infinite metric. In the case of RIP the infinite metric is 1.
Example: We have a network of two routers. Both routers are running RIP, R2 has advertised the 10.0.0.0/24 network to R1. Now consider what happens when the network 10.0.0.0/24 fails:
- R2 removes the route to 10.0.0.0/24 from its routing table.
- R2 advertises the 10.0.0.0/24 network with an infinitive metric (16) to R1 (“route poisoning”).
- R1 receives the update and knows that the network has failed. It will remove the route from its routing table.
