EIGRP Metric Calculation and Path Selection

1. The EIGRP Composite Metric

EIGRP is an advanced distance-vector protocol, but unlike RIP which simply counts hops, EIGRP uses a sophisticated composite metric to evaluate the quality of a path.

The potential components of this metric are:

• Bandwidth: The slowest bandwidth along the path.
• Delay: The cumulative delay along the path.
• Reliability: A measure of the link's stability.
• Load: How utilized the link is.
• MTU: The Maximum Transmission Unit of the path (used only as a tie-breaker, not in the calculation).

While EIGRP can use all these factors, by default, it only considers bandwidth and delay. This is a critical point to remember. These factors are weighted using constants called K-values.

The Classic Metric Formula

With the default K-values (K1=1, K3=1, K2=K4=K5=0), the complex formula simplifies significantly. The metric for a path is calculated as:

$\text{Metric}=\left(\frac{10^7}{\text{Slowest Bandwidth (kbps)}}+\frac{\text{Cumulative Delay (μs)}}{10}\right)\times 256$

Let's break this down:

• Bandwidth Component: EIGRP takes a reference bandwidth of 10 Gbps (10,000,000 kbps) and divides it by the slowest link's bandwidth in the path, expressed in kbps.
• Delay Component: EIGRP sums the delay values of all outgoing interfaces along the path. This delay is measured in microseconds (μs), but the formula uses it in tens of microseconds, hence the division by 10.
• Multiplier: The result is multiplied by 256 for backward compatibility with IGRP, EIGRP's predecessor.

To see a detailed walkthrough of this calculation, let's turn to a video explanation.

For a detailed textual reference and a table of default metrics for different interface types, the following Cisco Press article is an excellent resource.

2. Path Selection: Successors and Feasible Successors

Once a router calculates the metric for all possible paths to a destination, it needs to select the best one and identify potential backups. This is where the DUAL (Diffusing Update Algorithm) comes in, using the following key terms:

• Reported Distance (RD): The metric that a neighboring router reports to you for reaching a specific destination. It's the neighbor's own best metric (its Feasible Distance).
• Feasible Distance (FD): Your router's total metric to reach a destination through a specific neighbor. It is calculated as Your metric to the neighbor + that neighbor's Reported Distance.
• Successor: The path with the lowest Feasible Distance. This is the best, primary route, and it is installed in the routing table.
• Feasible Successor (FS): A backup path that is guaranteed to be loop-free. Having an FS allows EIGRP to converge almost instantly if the successor path fails.

The Feasibility Condition

For a path to qualify as a Feasible Successor, it must meet a critical rule known as the Feasibility Condition:

The Reported Distance (RD) of the potential backup path must be less than the Feasible Distance (FD) of the current successor path.

RD_backup < FD_successor

This rule is EIGRP's genius loop-prevention mechanism. If a neighbor's own metric to a destination (its RD) is already worse than your best total metric (your successor's FD), there's a chance that neighbor is actually using you to get there. Trying to use that neighbor as a backup would create a routing loop. By enforcing this condition, EIGRP ensures that any Feasible Successor is on a completely independent, loop-free path.

Let's watch a detailed explanation of this process.

3. Scaling for Speed: Wide Metrics

The classic EIGRP metric worked well for years. However, as network speeds soared past 1 Gigabit and into the 10G, 40G, and 100G range, a mathematical limitation emerged.

Take a look at Table 2-7 in the Cisco Press article you read. Notice that the metric for a 10 Gbps interface is very low (512). The formula simply doesn't have enough granularity to differentiate between a 10 Gbps link and a 40 Gbps link—they both calculate to the same minimum metric. EIGRP would see them as equally good, which is clearly not ideal.

To solve this, Cisco introduced EIGRP Wide Metrics, which are used in EIGRP Named Mode.

Key enhancements of Wide Metrics:

• Higher Granularity: The metric is based on a 64-bit calculation instead of 32-bit, providing a much wider range of values.
• Better Scaling: A larger scaling factor is used, allowing EIGRP to distinguish between very high-speed links.
• Finer Delay Measurement: Delay is calculated in picoseconds instead of microseconds.
• Future-Proofing: A new K6 value was added for future use, such as incorporating jitter or other link characteristics.

To understand the specifics of this evolution, let's return to the Cisco Press article.

You can verify which metric style a router is using with the show ip protocols command. If you see a K6 value listed in the output, the router is using Wide Metrics.

Conclusion

You have now analyzed the core mechanisms that drive EIGRP's path selection decisions. This knowledge is not just theoretical; it directly impacts network design, traffic engineering, and troubleshooting. When a path isn't behaving as you expect, the metric is one of the first places you'll need to look.

Key Takeaways:

• Classic Metric: By default, EIGRP calculates its metric using bandwidth and delay, weighted by K-values. The formula is (10^7/BW + Delay/10) * 256.
• Path Selection: The best path, or Successor, has the lowest Feasible Distance (FD).
• Feasibility Condition: A loop-free backup path, or Feasible Successor (FS), must have a Reported Distance (RD) that is strictly less than the FD of the successor. This is EIGRP's primary loop-avoidance mechanism.
• Wide Metrics: An enhancement to the classic metric, used in named mode, that provides the granularity needed to differentiate between modern high-speed links (above 1 Gbps). It uses a 64-bit calculation and measures delay in picoseconds.

Preview of the Next Lesson:

Now that we understand how EIGRP calculates metrics and chooses paths, our next step is to learn how we can influence this process. In the next lesson, we will configure and verify EIGRP route summarization at network boundaries. Summarization is a key tool for creating more stable and scalable routing domains by reducing the size of the routing table and limiting the scope of routing updates.