EIGRP feasible successor selection logic

[cosmingrosu]

I am reading a bit about EIGRP in the ENCOR study guide, and I found that for a route to be considered a backup route, the Reported Distance received for that route must be less than the Feasible Distance calculated locally. This logic guarantees a loop-free path.
As an example:
P 10.4.4.0/24, 1 successors, FD is 3328
        via 10.13.1.3 (3328/3072), GigabitEthernet0/1
        via 10.14.1.4 (5376/2816), GigabitEthernet0/2
where we can see that second route is backup because 2816 (Reported Distance) is lower than 3328 (Feasible Distance) of the first route.

My question is why is Reported Distance used as a factor in selecting a backup route? Why not use the actual Path Metric instead?
Is it possible for routing loops to occur if backup routes would be selected based on Path Metric instead of Reported Distance?

[David Ruess]

Correct. So the feasible distance is my neighbors reported distance to me plus MY link metric to get to that route. The reported distance is the metric for my neighbor to get to that route. So the logic is if the reported distance is less than the feasible distance the path can’t possibly go through me so it’s a loop free path.
 

As to your statement of why not use path metric instead. We are using path metric. It’s just the path metric as seen by the neighboring (next hop) router. The reported distance is the path metric up until that neighboring router.

I will also add that there are cases where the reported distance is more than the feasible distance and that route won’t become a feasible successor. However that doesn’t mean it automatically causes a loop. If metrics have been messed with or there is a longer loop free path since EIGRP is distance vector and only know what it’s neighbors report it can’t see that even though it’s a higher metric it’s still loop feee.

The feasibility condition GURANTEES loop free but every loop free path does not always meet the feasibility condition.

Hope that helps 

-David

[MHM Cisco World]

https://www.youtube.com/watch?v=TOzB1qIjxKI

this simple youtube is the best to explain the idea of FS, 
take look 

MHM

[M02@rt37]

Hello @cosmingrosu,

In EIGRP, the Reported Distance (RD) is not used to determine backup routes. The backup route selection in EIGRP is based on the Feasible Distance (FD) and Successor-Feasible Distance (SFD) values.

The FD represents the total metric to reach a destination as calculated locally, taking into account the metric of the outgoing interface and any additional costs configured on the router.

The SFD is the metric of the best path among all feasible successors. Feasible successors are alternative routes to the destination that have an advertised distance less than the FD of the current successor route.

When a route's FD is greater than the SFD of a feasible successor, the feasible successor becomes a candidate for being the successor route in case the current successor fails. This mechanism helps ensure loop-free path selection and provides backup routes that are ready to take over in case of primary route failure.

[David Ruess]

Not sure what you meant but the RD IS used in backup calculation. The RD is compared against the Feasible Successors FD. If its lower then the route is a Feasible Successor.

Formula:

The Reported Distance (RD) of the Feasible Successor (FS) must be lower than the Feasible Distance (FD) of the Successor

-David

[M02@rt37]

Hello @David Ruess,

Apologies for any confusion caused. In EIGRP, the Reported Distance (RD) of a Feasible Successor (FS) is compared to the Feasible Distance (FD) of the Successor to determine if the FS can be considered as a backup route.

To clarify, the Feasible Distance (FD) is the metric calculated by a router for a particular route based on its own local calculation, taking into account the bandwidth and delay of the path. The Reported Distance (RD) is the metric value advertised by a neighboring router for the same route.

When comparing the RD of an FS to the FD of the Successor, if the RD is lower than the FD, it means that the FS has a lower calculated metric than the current best route. In such a case, the FS can be considered as a backup route that can potentially replace the Successor in case of link failure, as it provides an alternative loop-free path to the same destination.

Using the RD and FD comparison helps ensure that backup routes are selected based on a lower metric value, which reduces the chances of routing loops. This mechanism helps maintain loop-free paths and provides resiliency in the network.

[Kanan Huseynli]

Hi,

below is what Jeff Doyle wrote in "Routing with TCP/IP Volume 1 (2005)":

The concepts of feasible successors and the FC are central to loop avoidance. Because feasible successors are always "downstream," (that is, a shorter metric distance to the destination than the FD) 
a router will never choose a path that will lead back through itself, creating a loop. Such a path would have a distance larger than the FD.

RD is best path metric with respect to neighbor, when this value (RD) is lower than local router FD it means that, FS (neighbor) does not use local router to reach the destination (if it would then RD should be higher than FD). And this neighbors FS (or path) can be used in case of primary path failure, for example because it is 100% loop free.

When RD is higher than FD then, local router knows that neighbor may use local router for the destination and path is not 100% loop-free (also, it is normal that RD can be higher than FD, but neighbor does not use local router. However, this can not be verified by local router). Thus this is not valid path until local router goes to ACTIVE and verifies (100% ensures) after DUAL is triggered due to some event.