EIGRP split horizon

[olly ahmed]

Hi,

I have a question regarding the split horizon. As per my reading I understand if a router learn a route via any of it's interface it does not send the route back using that same interface. I designed a simple EIGRP network with just 3 routers (R1,R2,R3), I found R1-R2, R2-R3 & R1-R3 are neighbors. I have configured a loopback interface on R2 and this route is being advertised by EIGRP. R1 and R3 learns this route by EIGRP from neighbor R2. But when I look into the database I found both R1 and R3 are leaning the same route from one another. My question is, if R1 learns about the route that R3 learns from R2 using a specific interface (fa1/0) why it is sending back the same route to R3 again using the same interface (fa1/0). For clear understanding I am attaching the snap and EIGRP topology output.

R1#show ip eigrp topology 5.5.5.5 255.255.255.255

IP-EIGRP (AS 1): Topology entry for 5.5.5.5/32
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 156160
Routing Descriptor Blocks:
2.2.2.1 (FastEthernet0/0), from 2.2.2.1, Send flag is 0x0
Composite metric is (156160/128256), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5100 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 1

3.3.3.1 (FastEthernet1/0), from 3.3.3.1, Send flag is 0x0
Composite metric is (158720/156160), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5200 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2

R3#show ip eigrp topology 5.5.5.5 255.255.255.255

IP-EIGRP (AS 1): Topology entry for 5.5.5.5/32
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 156160
Routing Descriptor Blocks:
1.1.1.1 (FastEthernet0/0), from 1.1.1.1, Send flag is 0x0
Composite metric is (156160/128256), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5100 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 1

3.3.3.2 (FastEthernet1/0), from 3.3.3.2, Send flag is 0x0
Composite metric is (158720/156160), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5200 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2

R2#show ip eigrp topology 5.5.5.5 255.255.255.255

IP-EIGRP (AS 1): Topology entry for 5.5.5.5/32
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 128256
Routing Descriptor Blocks:
0.0.0.0 (Loopback1), from Connected, Send flag is 0x0
Composite metric is (128256/0), Route is Internal
Vector metric:
Minimum bandwidth is 10000000 Kbit
Total delay is 5000 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1514
Hop count is 0

[Richard Burts]

It would have been helpful if your diagram had included identification of interfaces and interface IP addresses. But based on the output that you posted I see normal behavior of EIGRP.

R1 has learned 5.5.5.5 on interface FA0/0 from 2.2.2.1

R1 has not advertised it out the same interface but has advertised 5.5.5.5 out interface FA1/0

R3 has learned 5.5.5.5 on interface FA0/0 from 1.1.1.1

R3 has not advertised it out the same interface but has advertised 5.5.5.5 out interface FA1/0

So R1 has learned the best path to 5.5.5.5 via FA0/0 and has learned a second path via FA1/0 and there is no issue of split horizon in this.

So R3 has learned the best path to 5.5.5.5 via FA0/0 and has learned a second path via FA1/0 and there is not issue of split horizon in this.

So R1 has learned the best path from R2 on FA0/0 and has learned a backup path from R3 on FA1/0 which will provide redundancy if the best path fails. This is a feature of EIGRP and does not violate any part of split horizon. And the same thing applies to R3 which learns the best path from R2 and learns a backup path from R1.

HTH

Rick

[t.tineo]

ollyahmed,

Agree with Rick, the routers are learning and advertising the route via different interfaces.

What I find interesting is that R3 seems to be learning the best path via R1 instead of R2. 

From above:

R3#show ip eigrp topology 5.5.5.5 255.255.255.255

IP-EIGRP (AS 1): Topology entry for 5.5.5.5/32
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 156160
Routing Descriptor Blocks:
1.1.1.1 (FastEthernet0/0), from 1.1.1.1, Send flag is 0x0
Composite metric is (156160/128256), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5100 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 1

Guessing the following is the connection to R2. Delay is higher and has a hop count of 2. Guessing R2 & R3 don't have a direct connection. 

Rick mentions a diagram, but I don't see it. 

3.3.3.2 (FastEthernet1/0), from 3.3.3.2, Send flag is 0x0
Composite metric is (158720/156160), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 5200 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2

[t.tineo]

Saw the diagram. They are all supposed to be directly connected. The output doesn't match the diagram.