In order for a route to be considered a Feasible Successor and be INSTALLED in the Topology Table, the route's Advertised Distance MUST be LESS than the CURRENT Feasible Distance.
You may find the following useful in understanding the advertised values and the WHY rather than just having to accept that a router will choose the best value.
Yep. The RD is the distance the neighbor told you. The FD is when you also consider the link to get to the neighbor. The FD for RTRx should be > the RD received on a link by RTRx. Otherwise, you are going away from the destination. As another note, the FD for RTRx is the AD (advertised distance) sent to other routers further downstream. This AD is their RD and they calculate the full path for their FD.
Now, I live in Kentucky. I can drive from here to Orlando in about 10 hours. I could drive to Cincinnati first, but I hear that it is like 13 hours to Orlando from there. Therefore, that is not a feasible route.
So, for the above statement, i understand that if an EIGRP router advertises to me a composite metric that is lower than what i know to a particular destination network, then it guarantee's it lie's on a loop free route to the destination.
How and why does it guarantee this?
Here is how good folks explained it to me, so I'm eager to pass the word on
Consider the following topology.
R1 is our router which has a route to R5. Suppose the path through R2 is better (the lowest FD), so R2 is becoming our Successor for that route. Now R1 has to decide if R3 will become a Feasible Successor or not. Let's split this into 2 cases.
1) Suppose R1's distance to R5 (which is FD) is 100, whereas R3's distance to R5 through R4 is 120 (which is RD). Let's violate the feasibility condition and choose R3 as a Feasible Successor.
What happens when we configure unequal load balancing? R1 sends some IP packets to R5 through R3. When a packet comes to the next hop, R3 thinks "I need to send it towards R5, so let's consult my routing table". And guess what? The routing table will tell R3 to send the packet back to R1 because R1 is the Successor for that route, in R3's point of view (remember that distance on the left is 100+X, and 120 on the right; X is distance between R1 and R3 which may easily be less than 20). So we have a loop, as the packet can bounce between R1 and R3 never going to other routers.
2) Suppose R1's distance to R5 (which is FD) is 100, whereas R3's distance to R5 through R4 is 95 (which is RD). Now whatever distance X is between R1 and R3, 100+X always greater than 95, so R3 will never route through R1. This is why it's a loop free path.
Here is the concept of Eigrp FD and RD along with the Feasibility condition:-
1- The Reported Distance Or Advertised distance , is the metric to reported metric from the upstream router to the Destination. (In your example router D recieves two reported distance to reach router A, the first reported distance from router B which advertises the metric from router B to router A. The Second reported Distance from Router C which advertises the metric from Router C to Router A). So , the reported distance is the metric from the upstream router to the destination.
2- The Feasible Distance is simply , the total calculated metric to the destination. Looking at your example, the FD is the total metric from Router D to router A , So router D has two FD, one through router B and the other through Router C.
3- Now, your first question was if they are both the sme distanc e away? I would say no , because the AD is different and the FD of both paths are different , So they are not in the same distance away.
4- Your last question was, why wouldnt route through router C be a feasible successor? It would indeed meat the feasibility condition and its actually a "Feasible Successor".
Note to Remember:
The Feasibility Condition stays the following:
" In order for a route to become a feasible Successor, Its advertised distance (RD) must be lesser than the Feasible Distance (FD) of the Successor".
If you look at your example, you will find the path through C meets the feasinbility condition and becomes a "feasible successor" because its:
RD < FD of path B.
I hope I have clarified this for you,
A route is not a feasible successor if it does not meet the feasibility condition. This compares the RD of the FD of the local router. Your last statement says 'greater than the FD of the successor'. This is sort of circularly logic. By successor, we must be talking about a neighbor router. The FD of a neighbor router is it's ad or advertised distance. So that advertised distance becomes the local router's RD for that particular route/neighbor. It is this RD that is used in the feasibility calculation against the local FD. The FD would be the the best calculated local route and includes the link to get to the neighbor(s). It is this bandwidth/delay to get to each neighbor that is not included in the RD value. This can take some time to wrap your head around
Ok guys i think it all just clicked, i read all your posts, and just sat there with my brain hurting, what you all said made sence, but it just didn't click, but then all of a sudden, (like subnetting does!)
This is just for one destination network, i realise this happens for each and every network.....
"A router examines each and every possible route to a network by looking at the Reported Distance received from neighbouring routers to the particular network we are trying to reach, it then adds the distance from the reporting neighbour to the Reported Distance from the particular network, this gives the router a the feasible distance of each and every possible route to the destination network, it picks the route with the lowest Feasible Distance and stores it in the routing table as the Successor. It then looks for Feasible successors, so it examines all the other reported distances to the particular network in question, if the Reported Distance is greater than the original Feasible Distance, it can only be assumed that the router that advertised the Reported distance is going through the local router to get to that particular network, so it is not a Feasible Successor because it will cause a routing loop. If the Reported Distance is lower than the Feasible Distance, then it will be placed in the topology table as a Feasible Successor, this indicates it is one step closer to the destination than the local router"
Could i please ask you to correct me where i am wrong by correcting the above text and bolding your corrections......
Thankyou heaps, i love it when it clicks!!!!
Your understanding is correct, and regarding your first question, its true the router wont pickup the path through router C because it doesnt meet the feasibility condition.
If you examine the Topolgy indicated on the link above , you will find the RD of path C is greater than the FD of path B which is the successor, thats why it wasnt choosen ad a feasible successor.