Routing Decision

[M.Sultan]

Hello engineers

Look to below exhibition, OSPF,EIGRP,STATIC,RIP imagine any routing protocol is applied on routers for routing.

Now i intend to ping interface 10.10.10.10 /8 on R3, My default gateway is R1s interface IP, when packet arrives to R1, R1 forward packet to R4, how R4 understand which host ip i wanted to ping ? while we have the same interface ip on R2 on left side of diagram.

I know longest prefix match, Administrative distince, Metric, my question about router intellegince how it knows where to forward the packet i think router must be confused with this method ?

Thank you 

Diagram

[Flavio Miranda]

Hi,

  Based on the routes installed on the Routing table. when you setup interfaces with IP address and add a routing protocol, the IOS will biuld up , based on the information you informed, the routing table and the router will take decision based on that.

 And if you made a mistake and left no choice to the router, the route will not be installed on the routing table and the ping will fail with destination unrecheable message.

 On thing router never will be is confuse. They follow strict software rules and just that. Confuse is who configure the router on this situation.

[M.Sultan]

Thank you, you're confused with my question, ok imagine we installed OSPF, all routes advertised and full adjacency made, when i ping from Pc packets goes through my default gateway to R1 and then to R4, here how R4 knows which host ip we intended for ping ? while both R3 and R2 has network and host of 10.10.10.10/8 , Same prefix match, same AD, Same metric, however these parameters does'nt work until for same destination with different paths.

how R4 knows which Ip we want to ping ? 10.10.10.10/8 in R2 or 10.10.10.10/8 in R3 ?

Thank you 

[Flavio Miranda]

You are confused. Take a read about routing table, CEF  and RIB, understand better how a router works and you will answer you own question.

[MHM Cisco World]

depend 
if you use OSPF and 
A- equal cost to 10.x.x.x then the R4 will send half to R2 and half to R3
B- unequal cost to 10.x.x.x then the R4 will always use low cost path and send all traffic to R2 OR R3 
if you use EIGRP 
A- equal cost to 10.x.x.x then the R4 will send half to R2 and half to R3
B- unequal cost to 10.x.x.x then the R4 will always use low cost path and send all traffic to R2 OR R3, 
IF you config variance then the eigrp can do unequal cost and send again half to R2 and half to R3 

here you will face 50% drop if R4 use half 
here you will face 100 % drop or 100% success if R4 select right R(2or3)

[MHM Cisco World · Accepted solution]

I run lab to help you get idea 
R1 have two path to 10.0.0.0 one via R3 and other via R2 
the ping from R9 to R4 is drop because the R1 decided to forward traffic via R2 
why R1 decided that 
R1 have two path in RIB BUT still there is CEF that decided the egress traffic for specific traffic 
the longest match is not play role here because the R1 have two path same subnet mask /24 and hence both is same.

[Joseph W. Doherty]

Well for to begin with, having two interfaces with the same IP is, generally, a major error.

Next, you seem to assume, OSPF,EIGRP,STATIC,RIP all route the same way, they do not.  For example, RIP is Classful, i.e. to it, there's just 10.x.x.x network.  (I.e. same problem if your two interfaces has IPs of 10.1.1.1 and 10.2.2.2.)

Next, you can have the same destination be reachable via different paths.  How OSPF,EIGRP,STATIC,RIP deals with that also varies a bit based on protocol, including how the protocol is actually configurated.  

For example, many dynamic protocol will use multiple paths, to the same destination, if they "cost" the same to reach that destination.  However, such protocols will do that for a set number (often adjustable( like paths, but what happens if you exceed that limit, which path is chosen?

Routers won't be confused, but if you don't know all the "rules", you may certainly be confused by routing results.