Fine-Tuning EIGRP Metrics: Understanding Bandwidth, Delay, and K Values

April 2015   |   Reading Time: 10 min read

Introduction

Enhanced Interior Gateway Routing Protocol (EIGRP) remains a preferred choice in many enterprise environments for its balance of performance and simplicity. While often praised for its ease of deployment, EIGRP’s true power lies in its customizable metric system. Engineers who understand how bandwidth, delay, and K values interact can influence path selection with surgical precision. This post explores how to fine-tune EIGRP metrics effectively.

EIGRP Metric Formula

EIGRP uses a composite metric that includes bandwidth, delay, reliability, load, and MTU. The default metric calculation uses only bandwidth and delay, and the simplified formula looks like this:

metric = 256 * (K1 * bandwidth + K3 * delay)

By default, K1 and K3 are set to 1, and all other K values (K2, K4, K5) are set to 0. This means that only bandwidth and delay impact the routing decision.

Understanding Bandwidth in EIGRP

EIGRP treats bandwidth as the minimum bandwidth along the path. Lower bandwidth increases the metric and makes a path less preferred.

interface FastEthernet0/1
 bandwidth 10000

Note: This does not limit actual throughput. It's only used for metric calculations and QoS shaping.

Understanding Delay in EIGRP

Delay is cumulative and expressed in tens of microseconds. EIGRP adds the delay of each interface along the path. Lower delay means a better metric.

interface Serial0/0
 delay 2000

Be cautious when manually adjusting delay to influence routing—it impacts convergence and route selection.

Customizing K Values

You can modify which factors EIGRP uses by changing the K values, but this must be consistent across the entire EIGRP domain.

router eigrp 100
 metric weights 0 1 0 1 0 0

This configuration activates bandwidth (K1) and reliability (K4), which is uncommon and can introduce instability if not coordinated.

Why Not Use Reliability or Load?

• Reliability and load are dynamic and can fluctuate rapidly.
• Instability in these metrics can cause frequent route recalculations.
• Best practice is to avoid using K2, K4, and K5 in production unless explicitly required and monitored.

Influencing Routing Decisions

Adjusting interface delay is the safest and most predictable method to influence EIGRP metrics. Use it when you want to prioritize certain links over others without changing topology.

interface Serial0/1
 delay 1500

Metric Calculation Example

Assume a path with two interfaces: one with bandwidth of 100 Mbps and delay of 1000, the other with bandwidth of 10 Mbps and delay of 2000.

metric = 256 * (10^7 / min_bw + sum_delay)

metric = 256 * (10^7 / 10000 + 3000) = 256 * (1000 + 3000) = 1,024,000

Verification and Troubleshooting

• Use show ip protocols to verify active K values.
• Use show ip route eigrp to confirm path selections.
• Use show interfaces to check bandwidth and delay settings.
• Use debug eigrp fsm for deeper convergence analysis (use cautiously).

Best Practices

• Stick to default K values unless you have a clear use case.
• Adjust delay rather than bandwidth to influence metrics.
• Document any metric manipulations for audit and troubleshooting.
• Test changes in a lab environment before deploying to production.

Conclusion

Understanding how EIGRP calculates its composite metric gives network engineers a significant degree of control. Whether you're influencing primary paths or preparing for convergence scenarios, knowing when and how to tune bandwidth, delay, and K values can elevate your routing strategy. With great power comes great responsibility—so adjust carefully, document thoroughly, and always test before rollout.

Eduardo Wnorowski is a network infrastructure consultant and technologist.
With over 20 years of experience in IT and consulting, he brings deep expertise in networking, security, infrastructure, and transformation.
Connect on Linkedin