Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Understanding EIGRP

      In this articl eI’m going to talk about Cisco proprietary routing protocol, EIGRP or Enhanced Interior Gateway Routing Protocol. If you are familiar with RIP, which is Distance Vector protocol, EIGRP is classified as advanced Distance Vector protocol, you will find that most rules that are applied to RIP are also applied to EIGRP.

Let’s have a look at this table

How EIGRP forms neighbor relationship?

As a part of its operation, EIGRP forms neighbor relationships by sending Hello packets to other routers. If a router sends hello packets to another router that receives them,it will become a neighbor and they will just exchange their routing information that will be saved in the topology table.  The address used to advertise these ‘hello’packets is well-known, reserved multicast address

By default,EIGRP dictate hello and time values according to the network type.

Selecting the best path

EEIGRP is using a formula to calculate the metric which isreferred to as composite metric.

·        Bandwidth (K1)

·        Delay (K3)

·        Load (K2)

·        Reliability (K4)

Bandwidth and Delay are used bydefault and they have static values, for example FastEthernet link is 100 Mbitand has a delay of 100usec. Ethernet is 10 mbit and has a delay of 1000usec. Onthe other hand, Load and Reliability have dynamic values, because they changeover the time. Load, is how busy your link is and Reliability, is how yourreliable your link is. So, let’s have a look at the EIGRP formula

Metric = (K1 *bandwidth) +[(K2 * bandwidth / 256 – Load)] + (K3 * Delay

Or, let’s make itsimple

Metric =256*[10^7/BW(min) in Kbps] + [sum of delay in usec/10]

Where BW is the minimum bandwidth to reach adestination in Kilobits per seconds and Delay is in microseconds (usec)

In the above diagram, let’s calculatethe metric between R1 and network

As we can see, 64kbps is the minimumBW, and 2000 is the delay.

So, Metric = 256*[10^7/64] +[2000+2000+2000/10]

      Metric =256 * 156250 + 6000 =41536000 ß Bignumber, ha?

Forthe sake of simplicity, I will assign small values in our next diagram

Let’ssay that R4 is telling R3, it will cost me 15 to reach network call this Advertised Distance (AD). And R3 will say, to reach mydestination, network, it will cost 20 + 15 = 35. We call this asthe Feasible Distance (FD). In brief, Advertised Distance is the best metric a router receivesfrom the neighbor (next-hop device) to a given destination network/subnet, andFeasible Distance is the total metric value which is a sum of the AdvertisedDistance and the metric to reach the neighbor (next-hop router towards thedestination).

Well, can you tell me what is the ADand FD that are in the topology table of R1?

Let’s do it together,

R3 will tell R2 that, it will cost me 35 to reachto the destination network and R2 will know then that it will cost him 35 (ADof R3) + 15 (cost to reach R3) = 50 and this would be the FD for R2. But as R2is a neighbor to R1, it will tell R1 that it will cost 50 to reach thedestination network, and this will be recorded as the AD in R1 topology table,and sure R1 will add the cost for him to reach R2 which is 10, and this givesus the FD of R1 to reach the destination network that will be 60.

In the above diagram, we can see thatR1 has two paths to network So which path will R1 take to reachthat network? Obviously, the Ethernet link between R1 and R3 10 Mbit) is slowerthan the FastEthernet link between R1 and R2 (100 Mbit), this means that thebest path is R1-R2 and this is called “successor” and the other path, will bethe backup path, which is called “feasible successor”, but why it will be thefeasible successor??

To become the feasible successor,there is a formula you should know:

            Advertiseddistance of feasible successor < Feasible distance of successor

The feasible distance here between R1 andR2 is 15, while the advertised distance of R3 is 5, that’s why the link betweenR1 and R3 becomes the feasible successor.

Implementing EIGRP

We need two steps to implement EIGRP:

1.       To enable EIGRP process in the ‘config’mode.

2.       To instruct EIGRP which interfaces shouldparticipate in the EIGRP domain.

As soon as we enable EIGRP on aninterface, ‘hello’ packets will be sent from that interface using the multicast address to discover the neighbors. That interface will allow EIGRP process to read the address and network mask configured on this interface andadvertise them to the discovered neighbor(s).

The command that we use to enable EIGRPis: router eigrp 1

Number ‘1’ is the autonomous system number(between 1-65535) which must be the same on all EIGRP neighbors.

EIGRPis partially a Distance Vector routing protocol. By default the automatic summarization to the class boundary takes place. The command that disables that behavior is: 'no auto-summary' used under EIGRP process.

EIGRPuses five different types of the packets to communicate. Three of them are reliable and must be acknowledged by the receiver. The EIGRP packets are asfollows:

  1. Hello (unreliable).
  2. Updates (reliable).
  3. Queries (reliable).
  4. Replies (reliable).
  5. Acknowledgement (unreliable.

EIGRP has the ability to authenticate its packets, and currently (as the time of writing this article) uses only one protocol which is MD5. To configure EIGRP authentication, you have to two major steps that should be configured:

1.       Configuring a key chain with parameters such as password and the date/time.

2.       Applying authentication method and the key chain on per-interface basis.

Key authentication example:

R1(config)#key chain (keyname)

R1(config-keychain)#key 1


Note: ‘key 1’ – the number here must match theone that is used on the connected neighbor’s interface.

!enabling authentication on an interface

R1(config-if)#ip authenticationmode eigrp 1 md5

R1(config-if)#ip authenticationkey-chain eigrp 1 (key name)


Note: key-chain eigrp 1 – EIGRP ASnumber that is configured in the ‘config’ mode.

EIGRP Summarization

Let’s have a look at thefollowing topology:

Nowlet’s configure basic EIGRP configuration:

Cne(config)#routereigrp 1





maher(config)#routereigrp 1



Nowlet’s the routing table of maher router

maher#showip route eigrp

  subnetted, 2 subnets

D       [90/30720] via, 00:01:08,


D        [90/30720] via192.168.0.1, 00:01:08,


Let’suse ‘ip summary-address eigrp’ command on cne router


Cne(config-if)#ipsummary-address eigrp 1

Let’scheck routing table now on maher router

Maher#showip route eigrp

     is subnetted, 1subnets

D         [90/30720] via192.16.0.2. 00:06:49,


Thetwo networks disappear and a single entry will appear on router maher.

To read more about EIGRP, please visit

1-Introduction to EIGRP

2- Configuring EIGRP