Skip to content

EIGRP Basic Concepts

November 27, 2011
  • Prerequisites: IP subnetting and basic IP concepts

EIGRP is Cisco’s proprietary routing protocol. Initially Cisco developed IGRP but has since deprecated it in favor of EIGRP. EIGRP is intended to be a scalable, high-performance, high-feature routing protocol. This article intends to show the basics of how EIGRP works and what features it offers.

At the high-level, EIGRP is a distance vector protocol. However, it uses some link-state properties so Cisco calls it a hybrid protocol.

Neighbor Discovery/Recovery

Neighbor discovery and recovery are the first of four components of EIGRP. Neighbor discovery is responsible for finding directly attached EIGRP enabled routers and recovery handles unreachable neighbors. Neighbor discovery is achieved using hello packets. Responses to the hello packets instigate the route exchanges.

Reliable Transport Protocol

The second main component of EIGRP its reliable transport mechanisms. As reliable packets can be more resource intensive, reliability is only guaranteed in cases which it is required. To further enhance efficiency, both multicast and unicast packets are used.

DUAL Finite State Machine

Most of the compexity of EIGRP comes in the DUAL finite state machine which is tasked with all route computations.

Protocol Dependent Modules

Protocol dependent modules are abstraction layers allowing EIGRP to work with multiple layer 3 protocols.

EIGRP Convergence Process

When an EIGRP process comes online, the router will send out a hello multicast packet to IP address 244.0.0.10 with protocol type 88. If any EIGRP enabled routers receive this packet, it will respond with a hello packet and an EIGRP update packet. The update packet contains all EIGRP routes the router is aware of. Upon reception of the update packet, the first router will send an acknowledgement packet as well as its own update packet containing its EIGRP routes. As happened before, the second router sends an acknowledgement packet when the update packet is received.

EIGRP now monitors the links to make sure routes are active and working. The fastest method is to watch interface states. If the interfaces moves from up/up to anything else, EIGRP will know the link is down and change to a new route, also known as a feasible successor. Feasible successors are detailed later in this document. The up/up status may not always properly describe the status of a link. A 15 second Hold Timer is used to detect these errors. If a periodic hello packet isn’t received within the Hold Timer’s window, a new route is chosen. The operation of hello packets and the Hold Timer are similar to how Spanning Tree Protocol works with BPDUs.

If a router detects a route has went down and routes need to be recomputed, it sends a query packet to its neighbor. This recomputation has to happen because there is no feasible successor. The routers all communicate to each other whether they own a new, valid route. These messages are propogated to all routers via reply and query messages.

This is a very simple explanation of what happens during reconvergence of a route with EIGRP. It is my intention to do the following posts in the future explaining more basics of EIGRP.

  • EIGRP basic configuration
  • EIGRP route computation process
  • Packet-pick-apart for EIGRP route computation
Advertisements

From → Tutorial

One Comment

Trackbacks & Pingbacks

  1. Router | dmohankumar.com

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s