EIGRP Wide Metric Calculation

Q: How Do I Configure Bandwidth In EIGRP?

We can use the IP bandwidth-percent eigrp command to set this to a lower value to prevent EIGRP from flooding your interface(s). EIGRP traffic will then be queued and rate-limited by the router. On serial interfaces, Cisco router IOS will use a default bandwidth of 1544kbps.

The EIGRP Metric has confounded many network experts who are trying to understand the protocol. The goal of this post is to define the Eigrp metric and show how to Eigrp wide metric calculation work.

EIGRP is a Cisco-developed routing protocol that dates back to the 1980s. As a result, Cisco was the sole vendor with the authority to implement the protocol. Until Cisco published the specification as an IETF draught in 1998.

The ability to include many various factors while computing a route's cost, or metric, is one of EIGRP's key advantages. EIGRP is one of the few routing protocols that can calculate costs based on any combination of bandwidth, load, delay, and reliability.

What exactly are EIGRP Wide Metrics?

EIGRP calculates a composite cost metric based on bandwidth, latency, dependability, load, and K values. This metric has a flaw in that it does not scale for high-bandwidth links.

In EIGRP topologies, the EIGRP Wide Metrics feature provides 64-bit metric calculations and RIB scaling. Only EIGRP-named mode setups provide 64-bit computations. 32-bit computations are used in EIGRP classic mode settings.

This topic explains how wide metric computation works quite well.

How does EIGRP calculate the metric?

Any routing protocol relies on metric calculation to function properly. The measure for a path is calculated by EIGRP using many criteria. By default, bandwidth and latency are used in metric calculations, but they can also include interface load and reliability.

By default, EIGRP determines the optimum path to a remote network based on line bandwidth and delay. The ability to include many various factors while computing a route's cost, or metric, is one of EIGRP's key advantages.

EIGRP is one of the few routing protocols that can calculate costs based on any combination of bandwidth, load, delay, and reliability. The EIGRP metric calculation uses a delay value in tens of microseconds. Simply divide the DLY in the show interface command by 10 to get the Delay value.

Reported Distance

The metric describing the best route to access a network is known as the feasible distance (FD). In the routing table, that route will be shown. The metric given by an adjacent router for a particular route is known as reported distance (RD). In other words, it is the metric of the neighboring router's path to the network.

Bandwidth

In EIGRP, bandwidth is defined as the maximum amount of data that can be transmitted. To get the bandwidth value for the EIGRP metric, divide 10,000,000 by the bandwidth (in kbps) of the slowest link along the path to the destination network.

Delay is a cumulative measurement, unlike bandwidth, which denotes the "weakest link."

What is Delay in EIGRP

The overall delay in the path, measured in tens of microseconds (s), is called delay. The EIGRP formula is based on the IGRP metric formula, with the exception that the output is increased by 256 to make the metric 32 bits instead of 24 bits.

Reliability

A route's dependability metric is just a snapshot of the route's current reliability at the time it was last promoted. Remember that the EIGRP metric is retaken from the IGRP protocol, which was a timer-driven protocol. EIGRP, on the other hand, is an event-driven protocol.

EIGRP Load

Everyone studying EIGRP specifics is familiar with the "famous" composite metric formula, but the advice to keep the K values intact (or at least leave K2 and K5 at zero) or risk EIGRP's inability to adapt to changing load scenarios is rarely acknowledged.

IGRP, the forerunner to EIGRP, used the same vector metric and a very similar composite metric formula, but it was a real-time distance vector protocol (like RIP) that broadcast its routing information at regular intervals. It was fair to include interface load and dependability in the composite metric calculation because they were regularly conveyed throughout the network.

Formulas and K-Values

Each of these characteristics is controlled by a K-value. Each of these K-values allows one of the aforementioned attributes to be considered, as well as the scale on which the attribute is evaluated.

K1 = Bandwidth

K2 = Load

K3 = Delay

K4 & K5 = Reliability

Each of these numbers is employed in a Composite Metric calculation devised by EIGRP.

EIGRP Metric:

256 * { K1*BW + [(K2*BW)/(256-load)] + (K3*delay) } * { K5/(reliability+K4) }

The K values are numbers ranging from 0 to 255. You can customize each value based on what you want to factor into the cost calculation for each route.

You can set the relevant K-value to Zero in your routing domain if you don't want to examine one of the properties listed above. Set the relevant K-value to one if you want to consider an attribute.

You can scale how heavily a given number is regarded because the K-values can be any value between 0 and 255.

Set the K1 value to 2, and the K3 value to 1, for example, if you want Bandwidth to be deemed twice as significant as Delay. Set K1 to 2 and K3 to 3 if you want to consider Bandwidth and Delay in a 2:3 ratio.

This is why EIGRP's cost comparison is so flexible: you can choose which qualities to compare and how relevant each attribute is to your routing domain.

It should be emphasized, however, that two routers must have to match K-values before they may be EIGRP neighbors. This makes sense since if one router prioritizes Delay while the other prioritizes Bandwidth, they may argue about the optimum approach to a destination network.

Wide metrics use 64-bit values instead of 32-bit values. It works only in Named Mode not in Classic Mode because the K6 value is required in Wide metric calculation. K6 is used to measure Jitter, energy, or other future attributes. A simplified topology is shown here to show-

How does the Eigrp wide metric calculation work?

To check the K6 value, run the below command

#show ip protocols

Metric weight K1=1, K2=0, K3=1, K4=0, K5=0 K6=0

EIGRP Wide metric Formula:

Throughput = 65536 * 10^7 / minimum bandwidth across all links in the path

Latency = 65536 * total delay (in picoseconds) / 10^6

Composite metric = Throughput + Latency

Configuration for R1, R2 & R3

R1#conf t

R1(config)#interface f0/0

R1(config-if)#ip address 155.1.79.1 255.255.255.0

R1(config-if)#no shut

R1(config-if)#exit

Forming Neighborship:

R1#conf t

R1(config)#router eigrp MYKEY

R1(config-router)#ad

R1(config-router)#address-family ipv4 unicast autonomous-system 10

R1(config-router-af)#network 155.1.79.0 0.0.0.255

R1#sh ip eigrp topology 155.1.67.0/24

EIGRP-IPv4 VR(MYKEY) Topology Entry for AS(10)/ID(155.1.79.1) for 155.1.67.0/24

State is Passive, Query origin flag is 1, 1 Successor(s), FD is 19660800, RIB is 153600

Descriptor Blocks:

155.1.79.2 (FastEthernet0/0), from 155.1.79.2, Send flag is 0x0

Composite metric is (19660800/13107200), route is Internal

Vector metric:

Minimum bandwidth is 100000 Kbit

Total delay is 200000000 picoseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

Originating router is 155.1.79.2

R2 Configuration

R2#conf t

R2(config)#interface f0/0

R2(config-if)#ip address 155.1.79.2 255.255.255.0

R2(config-if)#no shut

R2(config-if)#exit

R2(config)#interface f2/0

R2(config-if)#ip address 155.1.67.1 255.255.255.0

R2(config-if)#no shut

R2(config-if)#exit

Forming Neighborship

R2(config)#router eigrp MYKEY

R2(config-router)#address-family ipv4 unicast autonomous-system 10

R2(config-router-af)#network 155.1.67.0 0.0.0.255

R3 Configuration

R3(config)#interface f0/0

R3(config-if)#ip address 155.1.67.2 255.255.255.0

R3(config-if)#no shut

R3(config-if)#exit

R3(config)#router eigrp MYKEY

R3(config-router)#address-family ipv4 unicast autonomous-system 10

R3(config-router-af)#network 155.1.67.0 0.0.0.255

Wild Metric Calculation for the route 155.1.67.0/24 from R1

Throughput = 65536 * 10^7 / 100000 = 6553600

Latency = 65536 *200000000 / 10^6 = 13107200

Wild Metric = Throughput + Latency = 6553600+13107200 = 19660800

(minimum bandwidth across all links in the path=100000 kbps, Total delay=200000000 picoseconds)

FAQs: EIGRP Wide Metric Calculation

Q: How do I configure bandwidth in EIGRP?

Ans: We can use the IP bandwidth-percent Eigrp command to set this to a lower value to prevent EIGRP from flooding your interface(s). EIGRP traffic will then be queued and rate-limited by the router. On serial interfaces, Cisco router IOS will use a default bandwidth of 1544kbps.

Q: What exactly is K6 in EIGRP?

Ans: K6 was introduced with Wide Metric support and is used to enable Extended Attributes, which can be used to reflect a higher aggregate metric than those with lower energy usage. In the scope of this document, there are currently two Extended Attributes defined: jitter and energy.

Q: What exactly is an offset list in EIGRP?

Ans: An offset list is a mechanism for increasing incoming and outgoing metrics for routes discovered using EIGRP or the Routing Information Protocol (RIP). Only distance vector routing protocols use offset lists. An offset list can optionally be limited by specifying an access list or an interface.

Q: Is load balancing enabled by default in EIGRP?

Ans: EIGRP by default supports equal-cost load balancing across four links. Equal-cost means that multiple routes must share the same metric in order for the router to load balance across equal-cost links.

That's all.

Explore More: