EIGRP Unequal Cost Routes

By Wendell October 6, 2015 09:05

#EIGRP uses a complex metric calculation. Most of us do not memorize the math formula, instead remembering the metric components: slowest bandwidth in the route and cumulative delay for all links. This lab lets you ponder how to configure EIGRP to use multiple unequal-metric routes, which begs many questions about what metrics EIGRP would calculate, and about how to configure correctly to make sure you use those multiple routes.



First, enable EIGRP on all routers and interfaces shown in the figure. The lab begins with the IPv4 addresses configured; your job is to add EIGRP configuration, with a few specific requirements.

Additionally, configure the network so that R4 will add routes for both the left route (through R2) and the right route (through R3) to the LAN off R1. To do so, rather than making you remember the math and predict the computed metric of each route, assume that the composite metric for the worse-metric route is less than 3X the metric for the best route.

The specific rules for this lab are:

  • Enable EIGRP on all routers
  • Use ASN 40
  • Allow unequal-cost load balancing for routes with metrics no more than 3X the metric of the best route
  • Do not use wildcard masks in the EIGRP configuration on routers R1, R2, and R3
  • Just for fun (no other reason), use a wildcard mask when enabling EIGRP on R4, such that if additional interfaces were added to R4 in the future, those interfaces would not automatically be enabled for EIGRP
  • Assume all IPv4 addresses implied in the figure are configured (per the initial configurations shown here)
  • Assume all interfaces in the figure are up and working
  • Use all defaults unless otherwise required to accomplish the lab requirements


Figure 1: Router Diamond with IP Subnets


Initial Configuration

Examples 1, 2, 3 and 4 show the beginning configuration state of R1, R2, R3 and R4.

Example 1: R1 Config


Example 2: R2 Config


Example 3: R3 Config


Example 4: R4 Config



Answer on Paper, or Maybe Test in Lab

Next, write your answer on paper. Or if you have some real gear, or other tools, configure the lab with those tools.

If you do try this lab beyond just writing the answers on paper or in a text editor, you have a great advantage for finding the actual metrics calculated for the alternate routes. I would suggest using the show ip eigrp topology command on R4, look at the entry for subnet, and find the metric for the successor (the best route) and the feasible successor (the next-best route). Then you can decide whether your configuration should or should not be correct for allowing both these routes into R4’s routing table. Ultimately, seeing two routes to on R4 (with the show ip route command) proves the lab worked as intended.


Do this Lab with Cisco’s VIRL

You can do these labs on paper and still get a lot out of the lab. As an extra help, we have added files for the Virtual Internet Routing Lab (VIRL) software as well. The .VIRL file found here is a file that when used with VIRL will load a lab topology similar to this lab’s topology, with the initial configuration shown in the lab as well. This section lists any differences between the lab exercise and the .VIRL file’s topology and configuration.

Download this lab’s VIRL file!


Network Device Info

The .VIRL topology matches this lab topology exactly. The host info does as well.


Host Device Info:

This lab does not use any hosts.

Answers: EIGRP Unequal Cost Routes
Answers: OSPF Networking Config 1
By Wendell October 6, 2015 09:05
Write a comment

1 Comment

  1. Bojan October 6, 16:33

    So I’m testing this in the CCNA version of Boson and it doesn’t allow for the variance command (CCNP only!). I got as far as establishing the EIGRP and R4 has two routes as you said:

    State is Passive, Query origin flag is 1, 2 Successor(s), FD is 33280 (FastEthernet2/0), from, Send flag is 0x0
    Composite metric is (33280/30720), Route is Internal
    Vector metric:
    Minimum bandwidth is 100000 Kbit
    Total delay is 300 microseconds
    Reliability is 255/255
    Load is 1/255
    Minimum MTU is 1514
    Hop count is 2 (FastEthernet1/0), from, Send flag is 0x0
    Composite metric is (33280/30720), Route is Internal
    Vector metric:
    Minimum bandwidth is 100000 Kbit
    Total delay is 300 microseconds
    Reliability is 255/255
    Load is 1/255
    Minimum MTU is 1514
    Hop count is 2

    I think the only other thing required to make the lab work is “variance 3” in each router eigrp 40 section. I don’t think maximum-path needs to be adjusted as you didn’t indicate this aspect in your lab.

    On R1-3 I used a wildcard-less network statement for /24s, but on R4 I used a host specific network statement:

    router eigrp 40

    Devices on R1’s LAN can ping devices on R4’s LAN ( to

    Reply to this comment
View comments

Write a comment

Comment; Identify w/ Social Media or Email


Subscribe to our mailing list and get interesting stuff and updates to your email inbox.

Thank you for subscribing.

Something went wrong.