Welcome back!
In our last lesson, we explored the five EIGRP packet types, which are the fundamental communication tools of the protocol. We saw that when a primary route fails and there is no pre-calculated backup, a router sends Query and Reply packets to find a new path. This hints at a deeper intelligence at work.
Today, we will dissect that intelligence: the Diffusing Update Algorithm (DUAL). This algorithm is the engine that drives EIGRP, making it remarkably fast and ensuring a loop-free topology. Your goal for this lesson is to analyze the DUAL algorithm, including feasible distance, reported distance, and successor selection.
Understanding DUAL is not just an academic exercise; it's the key to diagnosing EIGRP behavior and designing resilient networks. For a solutions engineer, mastering this topic is essential for troubleshooting complex routing issues that go beyond basic configuration.
Let's begin by understanding the "currency" DUAL uses to make decisions: the EIGRP metric.
1. The EIGRP Metric: DUAL's Decision-Making Currency
Before DUAL can choose the "best" path, it needs a way to measure and compare paths. This is done using a composite metric. While the formula can look intimidating and involves several components (Bandwidth, Delay, Reliability, Load, MTU), in virtually all real-world networks, only two are used by default.
To start, let's watch a section of the "EIGRP Deep Dive" video by Kevin Wallace. He provides an excellent breakdown of the EIGRP metric, explaining which components are used by default and how the calculation works in practice.
Please watch from 17:57 to 25:42. Focus on these key points: The five components of the metric (Bandwidth, Delay, Reliability, Load, MTU). Which two components are used by default. How the bandwidth and delay values are derived and used in the simplified formula.
As you saw, the default EIGRP metric calculation is simplified to:
- Bandwidth: This is the slowest link bandwidth along the entire path to the destination, measured in kilobits per second.
- Delay: This is the sum of the delay values for all outgoing interfaces along the path, measured in tens of microseconds.
It's crucial to remember that by default, EIGRP only considers bandwidth and delay. This is controlled by constants called K-values. K1 (for bandwidth) and K3 (for delay) are set to 1 by default, while the others are 0. Changing these is possible but rarely done, as it requires identical configuration across all routers in the autonomous system to maintain neighborship.
2. The Core of DUAL: Successors, FD, and RD
With an understanding of the metric, we can now explore the three core concepts of DUAL. These terms define how a router perceives and ranks the paths it learns from its neighbors.
- Feasible Distance (FD): The local router's best calculated metric to a destination. Think of it as "my total cost to get there."
- Reported Distance (RD): The metric to a destination as advertised by a neighbor. This is the neighbor's own Feasible Distance. Think of it as "what my neighbor told me their cost is."
- Successor: The neighboring router that offers the path with the lowest Feasible Distance. This is the best, primary route that is installed in the routing table.
The distinction between FD and RD is all about perspective.
The concept of perspective can be tricky. This short video from Johnny Mac uses a simple but effective analogy to clarify the difference between Feasible Distance and Reported Distance.
Watch from 00:23 to 03:38. Pay attention to the 'Dad vs. Jamie' analogy and how the same value is considered an FD from one router's perspective and an RD from another's.
Now, let's see how these concepts are used to select the best path.
Let's return to Kevin Wallace's "EIGRP Deep Dive" to see a practical example of how a router calculates the FD for multiple paths and chooses a Successor.
Watch from 25:42 to 28:55. Observe how the router on the right (R3) calculates the FD for paths via R1 and R2 by adding its own cost to the RD from each neighbor. The path with the lower total FD becomes the Successor route.
3. Loop-Free Backups: The Feasible Successor
Having a fast, pre-calculated backup path is what gives EIGRP its rapid convergence. This backup is called a Feasible Successor (FS). However, not just any alternate path can be a feasible successor. It must pass a strict test to ensure it is 100% loop-free. This test is the Feasibility Condition.
Feasibility Condition: A path through a neighbor is a valid Feasible Successor if the neighbor's Reported Distance (RD) is strictly less than the local router's Feasible Distance (FD) for the current successor route.
RD_potential_backup < FD_current_best_path
Why this rule? If a potential backup neighbor has an RD that is greater than or equal to my own best metric (my FD), it's possible that this neighbor's path to the destination actually loops back through me. The feasibility condition is a conservative, mathematical guarantee to avoid this.
The following video provides a detailed walkthrough of this critical concept.
Kevin Wallace's video provides one of the clearest explanations available for the Feasibility Condition. He illustrates not only how it works but, more importantly, why it's necessary to prevent routing loops.
Watch from 28:55 to 35:56. This is a crucial section. Focus on understanding the logic behind the rule and how to apply it to the given network diagram to determine which paths qualify as Feasible Successors.
A path that meets the Feasibility Condition is stored in the EIGRP topology table as a backup. If the successor route fails, the router can instantly switch to the feasible successor without any recalculation or querying, leading to near-instantaneous convergence.
Test your understanding!
A router has three possible paths to the 172.16.10.0/24 network:
- Path via R1: Feasible Distance (FD) = 2000, Reported Distance (RD) = 1500
- Path via R2: Feasible Distance (FD) = 3000, Reported Distance (RD) = 1900
- Path via R3: Feasible Distance (FD) = 3500, Reported Distance (RD) = 2100
- Which path will be the Successor and why?
- Which path(s), if any, will be chosen as a Feasible Successor? Explain your reasoning using the Feasibility Condition.
Show answer
-
The path via R1 will be the Successor because it has the lowest Feasible Distance (FD) of 2000. This is the best metric from the local router's perspective.
-
The path via R2 will be the Feasible Successor.
- To check, we compare its Reported Distance (1900) to the Successor's Feasible Distance (2000).
- Since
1900 < 2000, the Feasibility Condition is met. - The path via R3 will not be a Feasible Successor because its Reported Distance (2100) is not less than the Successor's Feasible Distance (2000).
4. Viewing DUAL in Action: The Topology Table
All of this information—Successors, Feasible Successors, and even non-qualifying paths—is stored in the EIGRP topology table. This table is the complete database DUAL uses for its calculations. You can view it with the show ip eigrp topology command. The routes that are selected as Successors are then installed into the main IP routing table.
Understand and Use the Enhanced Interior Gateway Routing Protocol
The official Cisco documentation provides a detailed breakdown of the show ip eigrp topology command output. This will help you translate the theory of FD and RD into the practical information you see on a router's CLI.
First, quickly read the section "Feasible Distance, Reported Distance, and Feasible Successor" to see Cisco's formal definitions. Then, scroll down to the section "show ip eigrp topology " and review the "Configuration Explanations." Pay special attention to how the output line Composite metric is (307200/281600) maps directly to (Feasible Distance / Reported Distance).
5. What If There's No Feasible Successor?
As we discussed in the last lesson, if a Successor fails and there is no Feasible Successor in the topology table, the router doesn't give up. Instead, the route state changes from Passive (stable) to Active (searching).
The router then multicasts a Query packet to all its neighbors, effectively asking, "Does anyone have a loop-free path for this lost network?" The neighbors respond with Reply packets, and DUAL uses this information to find a new Successor. This process is slower than failing over to a Feasible Successor but ensures the network converges correctly and without loops.
Conclusion
In this lesson, we demystified the DUAL algorithm, the intelligent core of EIGRP. You now have a solid theoretical foundation for how EIGRP makes its path-selection decisions, which is crucial for advanced troubleshooting and network design.
Key Takeaways:
- DUAL's decisions are based on a metric that, by default, uses Bandwidth and Delay.
- Feasible Distance (FD) is the router's own best calculated metric to a destination. The path with the lowest FD becomes the Successor.
- Reported Distance (RD) is the metric reported by a neighbor. It is the neighbor's own FD.
- A loop-free backup route, or Feasible Successor (FS), is selected only if it meets the Feasibility Condition: its RD must be strictly less than the current Successor's FD.
- Having an FS allows for near-instantaneous convergence. If no FS exists, the router goes Active and uses the Query/Reply process to find a new path.
Preview of the Next Lesson:
We have now covered the "what" and the "how" of EIGRP's fundamental operations. In the next lesson, we will shift to modern, practical implementation. We will learn to configure EIGRP named mode with address families for both IPv4 and IPv6, which is the current best practice for deploying EIGRP in today's networks.