EIGRP variance over 2 routers?? Possible?

mattsthe2

I tried labbling this up in GNS3 no luck.


R1

---

WAN

---

R2
|......................|
R3

---

WAN

---

R4


Varience was placed on R3 an and R1. (I set a varience of like 80 so i made sure my metrics was ok)

If at R3 contacting a Loopback on R2 i wanted load-balancing routes in the table so it would have two paths:

Path 1: R3-R1-R2
Path 2: R3-R4-R2


For the life of me i cant get this working.


However if i terminate the two WAN Circuits into R3 it works fine. Like this:

R3---WAN---R2
..................|
|

---

WAN--R4

networker050184

A route has to be a feasible successor to be included in unequal cost load balancing. Might want to double check that.

mattsthe2

networker050184 wrote: »

A route has to be a feasible successor to be included in unequal cost load balancing. Might want to double check that.

Well it shows in the topology table. Does anyone know why?

R1#sh ip eigrp topology 172.17.162.1/32
IP-EIGRP (AS 100): Topology entry for 172.17.162.1/32
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 25600512
Routing Descriptor Blocks:
172.17.50.2 (GigabitEthernet1/0), from 172.17.50.2, Send flag is 0x0
Composite metric is (25600512/25600256), Route is External
Vector metric:
Minimum bandwidth is 100 Kbit
Total delay is 20 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 1
External data:
Originating router is 192.168.2.1
AS number of route is 1
External protocol is OSPF, external metric is 3
Administrator tag is 0 (0x00000000)
192.168.1.2 (GigabitEthernet2/0), from 192.168.1.2, Send flag is 0x0
Composite metric is (25600768/25600512), Route is External
Vector metric:
Minimum bandwidth is 100 Kbit
Total delay is 30 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2
External data:
Originating router is 192.168.2.2
AS number of route is 1
External protocol is OSPF, external metric is 2
Administrator tag is 0 (0x00000000)

networker050184

The feasible distance (FD) for the route is 25600512. Since the AD for the other route (25600512 also) is not lower than the FD it will not be used as a feasible successor.

If you do a show ip eigrp topology with the prefix it shows all paths not just successors and feasible successors.

mattsthe2

networker050184 wrote: »

The feasible distance (FD) for the route is 25600512. Since the AD for the other route (25600512 also) is not lower than the FD it will not be used as a feasible successor.

If you do a show ip eigrp topology with the prefix it shows all paths not just successors and feasible successors.

Excellant description. Totally get why its not a feasiblee successor. However 2 things, if i may.

1) my above output was a 'sh ip eigrp topology with the prefix'
2) I've attached a image showing the topology.
Basically R1 to R3 is a WAN link, as well as R2 to R4.
Is there any way to make that other route a feasible successor, as you can see there is no loops so why cant it be used?
Can i tweak the metrics some how to fool EIGRP into changing the AD
One other thing to note, R3 is running EIGRP and OPSF as well as R4. I have mutual redistrubtion going on.

R3#router eigrp 100
redistribute ospf 1 metric 100 1 255 1 1500 route-map NETS-OSPF-to-EIGRP
network 172.17.50.0 0.0.0.255
no auto-summary

R4#router eigrp 100
redistribute ospf 1 metric 100 1 255 1 1500 route-map NETS-OSPF-to-EIGRP
network 172.17.51.0 0.0.0.255
no auto-summary

thanks!!!!

networker050184

Yes you can tweak the metrics. Its easiest to change delay with EIGRP to get your traffic going where you want it.

Just curious why you want to do this? Seems like you are making something more complicated than it needs to be.

mattsthe2

There is a method to my madness its called my boss.
My lab is similar to one of our locations and the "WAN" links are 100Mb circuits. My boss wants me to investigate how to utilize both circuits as best as possible, without OeR. With the way our production traffic is routing its preferring one circuit over the other. I'm tasked with trying to utilize both circuits.

I'll mess with delay tomorrow when im at work and let you know.

As i said before when both "WAN" circuits term to the same device getting this to Load-Share should be a snap, but i cant get it do load share on separate devices and i get the output you have already seen.

Thanks for your help though!!!!

networker050184

The easiest way to do this is to put one device behind the wan routers and have two identical paths, one from each router.

Is one of the two wan links over subscribed? Why does he want load sharing? I don't understand why so many people want traffic to be so even. As long as there is an alternate path and either circuit can handle the full load it really doesn't matter. I know its your boss, but you guys are just making things harder on yourself by over complicating it....

mattsthe2

Well the delay thing worked out perfect, changing the delay with the varience added the Feasible Successor. Thanks for your help Networker
However i can't do it on R2 because of loop prevention.

Your suggestion is to put a device behind the WAN routers is a good one however i might suggest the reverse. Currently the EDGE routers are really 3750 switches which is also hosting different VLANs. I'm thinking about putting one device for the EDGE an actual router, that way it can do the load balancing and then just keep the 3750's for hosted VLANS.

burbankmarc

If you put a router in then you'll have a single point of failure. You could just stack the 2 3750's.

mattsthe2

Awesome idea!!
Thanks!