Eigrp Uses The Bellman-ford Algorithm For Calculating Its Metric

Calculate the EIGRP Composite Metric

This calculator determines the EIGRP classic composite metric based on the two default components: minimum bandwidth and total delay. EIGRP uses these values in a formula derived from the Bellman-Ford algorithm’s principles to find the best path in a network.

Metric Component Analysis

Example EIGRP Metric Values for Common Link Types

Interface Type	Bandwidth (kbps)	Default Delay (µs)	Calculated EIGRP Metric
Fast Ethernet	100,000	100	28160
Gigabit Ethernet	1,000,000	10	5120
10 Gigabit Ethernet	10,000,000	10	512
T1 Serial	1,544	20,000	2170104
56k Leased Line	56	20,000	46295056

This table shows the default EIGRP Metric for single-hop paths over various common network interface types.

What is the EIGRP Metric?

The EIGRP Metric is a composite numerical value used by Cisco’s Enhanced Interior Gateway Routing Protocol (EIGRP) to determine the best path to a destination network. While EIGRP is a distance-vector protocol, its path selection is based on a sophisticated metric, not just hop count. The core path-finding logic is handled by the Diffusing Update Algorithm (DUAL), which relies on the principles of the Bellman-Ford algorithm to ensure loop-free paths. This calculator focuses on the classic EIGRP Metric calculation, which by default, considers only bandwidth and delay. A lower calculated EIGRP Metric signifies a more preferred route.

Network administrators should use the EIGRP Metric to understand and influence routing decisions. By adjusting the bandwidth and delay parameters on router interfaces, you can control which paths are chosen. A common misconception is that the metric reflects real-time link performance; however, it’s based on static values configured on the interfaces, making the EIGRP Metric a predictable and stable path selection tool.

EIGRP Metric Formula and Mathematical Explanation

The power of the EIGRP Metric comes from its formula, which combines path attributes into a single comparable number. The default formula for the classic EIGRP Metric is:

Metric = 256 * (Scaled_Bandwidth + Scaled_Delay)

Where:
– Scaled_Bandwidth = 10⁷ / Minimum_Path_Bandwidth_kbps
– Scaled_Delay = Total_Path_Delay_µs / 10

The formula first scales the two key variables. Bandwidth is inverted—a higher bandwidth results in a lower scaled value, making it more desirable. Delay is summed up along the path and then scaled down. These two components are added together, and the entire result is multiplied by 256 to increase the granularity of the metric, a holdover from its predecessor, IGRP. This process ensures both the path’s speed (bandwidth) and its latency (delay) are factored into the final EIGRP Metric.

Variables Table

Variable	Meaning	Unit	Typical Range
Minimum Path Bandwidth	The slowest link’s bandwidth along the entire path to the destination.	kbps	64 – 10,000,000+
Total Path Delay	The cumulative delay of all outgoing interfaces along the path.	microseconds (µs)	10 – 100,000+
Scaled Bandwidth	The reference bandwidth (10 Gbps) divided by the minimum path bandwidth.	(unitless)	1 – 156,250
Scaled Delay	The total delay divided by 10.	(unitless, tens of µs)	1 – 10,000+
EIGRP Metric	The final composite value used for path comparison.	(unitless)	256 – Infinity

Practical Examples (Real-World Use Cases)

Example 1: Path over Fast Ethernet

Imagine a path to a server that traverses two Fast Ethernet links (100,000 kbps each). The delay for a Fast Ethernet interface is typically 100 microseconds.

• Inputs:
  • Minimum Bandwidth: 100,000 kbps (since both links are the same speed)
  • Total Delay: 100 + 100 = 200 µs
• Calculation:
  • Scaled Bandwidth = 10,000,000 / 100,000 = 100
  • Scaled Delay = 200 / 10 = 20
  • EIGRP Metric = 256 * (100 + 20) = 256 * 120 = 30,720
• Interpretation: The final EIGRP Metric for this path is 30,720. A router will compare this value against the metric of alternative paths to make its routing decision.

Example 2: Path over a Slower WAN Link

Now, consider an alternative path through a T1 serial link (1,544 kbps, 20,000 µs delay) followed by a Fast Ethernet link (100,000 kbps, 100 µs delay).

• Inputs:
  • Minimum Bandwidth: 1,544 kbps (the slowest link in the path)
  • Total Delay: 20,000 + 100 = 20,100 µs
• Calculation:
  • Scaled Bandwidth = 10,000,000 / 1,544 ≈ 6476
  • Scaled Delay = 20,100 / 10 = 2010
  • EIGRP Metric = 256 * (6476 + 2010) = 256 * 8486 = 2,172,416
• Interpretation: The EIGRP Metric is significantly higher (worse) than the first example, so EIGRP will strongly prefer the path over Fast Ethernet. To learn more about path selection, see our guide on Understanding Routing Protocols.

How to Use This EIGRP Metric Calculator

Using this calculator is a straightforward process to determine a path’s EIGRP Metric.

1. Enter Minimum Bandwidth: Identify the slowest link in the entire path from the source router to the destination network. Enter this value in kilobits per second (kbps) into the first field. You can find this using the `show interface` command on Cisco routers. For more details on commands, check our Common Cisco IOS Commands resource.
2. Enter Total Delay: Sum the delay values of all outgoing interfaces along the path. Enter this total in microseconds (µs) into the second field.
3. Read the Results: The calculator will instantly update, showing you the final EIGRP Metric. A lower metric is always better. The intermediate values show the individual contributions of bandwidth and delay to the final score.
4. Analyze and Decide: Use the result to predict routing behavior or to plan network changes. If you want to influence EIGRP to choose a different path, you can adjust the interface bandwidth or delay values on your routers to change the EIGRP Metric.

Key Factors That Affect EIGRP Metric Results

Several factors influence the final EIGRP Metric. Understanding them is key to network engineering.

• Bandwidth: This is the most significant factor. The metric is inversely proportional to the minimum bandwidth on the path. Halving the bandwidth will roughly double the bandwidth component of the metric.
• Delay: This is an additive metric that represents latency. It’s summed across every hop. Paths with more routers will have a higher total delay and thus a higher EIGRP Metric. For an in-depth look, see our article on Network Latency Explained.
• Hop Count: While not a direct factor, hop count implicitly affects the metric by increasing the total delay, as each hop adds its interface’s delay to the sum.
• K-Values: This calculator uses the default K-values (K1=1, K3=1, others=0), which only consider bandwidth and delay. If K-values are changed to include reliability or load, the EIGRP Metric calculation becomes more complex and dynamic, which is generally not recommended.
• Reference Bandwidth (10⁷): The formula uses a reference bandwidth of 10 Gbps (10,000,000 kbps). This means that links faster than 10 Gbps might not be differentiated effectively using the classic EIGRP metric, leading to potentially suboptimal routing. This is addressed by EIGRP Wide Metrics.
• Administrative Configuration: A network administrator can manually set the bandwidth and delay on an interface to influence the EIGRP Metric. This does not change the link’s actual speed but fools EIGRP into seeing the path as more or less desirable.

Frequently Asked Questions (FAQ)

1. Why is the EIGRP Metric multiplied by 256?

This is for backward compatibility with EIGRP’s predecessor, IGRP (Interior Gateway Routing Protocol). IGRP had a 24-bit metric, while EIGRP uses a 32-bit metric. The multiplication scales the value to fit the larger field.

2. What algorithm does EIGRP use for path calculation?

EIGRP uses the Diffusing Update Algorithm (DUAL). DUAL is a network discovery and path-finding algorithm that uses the EIGRP Metric as its cost value. It’s based on the principles of the classical Bellman-Ford algorithm to guarantee loop-free routing.

3. Is a higher or lower EIGRP Metric better?

A lower EIGRP Metric is always better. Routers will install the path with the lowest metric into the routing table as the primary route (the successor).

4. What are the K-values in the full EIGRP metric formula?

The full formula includes five constants, or K-values (K1 to K5), that act as toggles or multipliers for bandwidth, load, delay, and reliability. By default, only K1 (bandwidth) and K3 (delay) are set to 1, disabling load and reliability from the calculation. If you’re comparing routing protocols, you might be interested in our OSPF Cost Calculator as well.

5. Does changing the interface bandwidth affect the link speed?

No. Using the `bandwidth` command in the interface configuration on a Cisco router only changes the value used for routing protocol calculations like the EIGRP Metric. It does not alter the actual physical speed of the link.

6. Why don’t we use load and reliability by default?

Load and reliability are highly dynamic values that can change frequently. Including them in the EIGRP Metric calculation would cause routers to constantly recalculate paths and send updates, leading to network instability. Static values like bandwidth and delay provide stable, predictable routing.

7. How does the Bellman-Ford algorithm relate to EIGRP?

EIGRP’s DUAL algorithm is a distance-vector algorithm that evolved from the concepts pioneered by Bellman-Ford. It exchanges vectors of distances (metrics) with neighbors to build a view of the network topology and uses this information to calculate the lowest-cost, loop-free path to every destination.

8. Where can I learn more about other routing protocols?

For a deep dive into another popular protocol, check out our comprehensive guide on BGP Path Selection.

Related Tools and Internal Resources

Explore other networking tools and resources to expand your knowledge.