EIGRP Distance Vector Route Discovery

April 27, 2014

EIGRP’s route discovery process uses a “routing by rumour” approach. Meaning that the routers in an autonomous system heard about their routes from another autonomous system/router, they did not receive the route information first hand.This essentially means the routers heard about the routes by listening to network gossip.

The information collected is stored in the following tables.

  • Neighbour Table
  • Topology Table
  • Router Table

With this EIGRP will calculate the best paths and any redundant links (if available) using the following:

  • Bandwidth
  • Delay
  • Load
  • Reliability

MTU is sometimes reffered to being used but it in fact is not, it is only used for some commands but doesn’t affect the outcome of the formula.


EIGRP and DUAL

April 19, 2014

DUAL or Diffusing Update Algorithm is responsible for the following.

 

  • Responsible for selecting and maintaing the best path to remote networks
  • DUAL also allows EIGRP to execute the following:
    • Calculate backup routes (if available)
    • Allows support of VLSM
    • Allows EIGRP to perform Dynamic Route Recovery
    • Questions EIGRP neighbours for any unknown alternate routes
    • Queries for alternate routes
    • Is the reason EIGRP can converge amazingly fast.

Reliable Transport Protocol (RTP)

April 16, 2014

This is a brief rundown of RTP and how it is used by EIGRP for reliable communication.

  • Proprietary Protocol
  • Used and relied on by EIGRP for reliable packet delivery to neighbours.
  • Manages inter EIGRP router communications

So EIGRP uses the multicast address 224.0.0.10 and each time a multicast is sent, it is recorded and awaits a reply of receipt. If and when the multicasts fails, EIGRP will switch to unicast packets and continue to await a reply. If there is no response after 16 attempts, that neighbour is declared dead.

So any information these routers send between neighbours is tracked and tagged with sequence numbers, this allows the routers to identify:

  • Old
  • Out of order
  • Redundant Information

EIGRP Metric Terminology

April 15, 2014

Feasible Distance:

One of the metrics used to calculate the best path and it is the best cost path as determined by the “Local” router

Reported Distance:

Reported or Advertised Distance is a metric that is calculated by a remote neighbour in a remote network.

 

Nb. The route with the lowed FD is the route that will populate the EIGRP routing table.

Neighbour Table:

Stored in RAM, this table is a record of all adjacent neighbours address and interfaces. Each router will use this table to track the current state information of its neighbours

Topology Table:

This table records destinations as reachable by its neighbours or simply the topology of the network. Any changes will be triggered with partial updates

Routing Table:

Using the information found in the topology table, EIGRP will place the best routes into the routing table. known as successor routes, they provide the least cost to reach the remote network.

Feasible Successor:

An inferior alternate route to the successor, which is located in both the topology and routing table, whereas the feasible successor is only recorded inside the topology table. If the successor route was to fail this would be considered the backup route and take over.

It is good to note that in IOS 15.0 code you can keep upto 32 feasible successor routes in the topology table and previous codes upto 16.

Successor Route:

This is the best route to a remote network, and as mentioned previously it is recorded in both the routing and topology tables.


Distance Vector and Routing Loop Preventions

April 2, 2014

Distance vector routing protocols are considered the most efficient routing protocols when it comes to the use of the device resources. Also unlike Link State protocols, who will share the entire routing “map” with every other adjacent device, Link State protocols prefer to share only the bare essentials and advertise only the best route.

To compare, Link Sate will advertise all routes to a destination but in Distance Vector they will only advertise the “best”. This helps keep the routing tables small and like i said at the  start, the router operates more efficiently and uses less device resources.

Distance Vector also has some route looping mitigation techniques in operation, to help stop Routing Loops from forming.

They are:

  • Triggered Updates
  • Route Poisoning
  • Maximum Distance (Hops)
  • Split Horizon
  • Hold Down Timers

These are all on and running by default and usually they won’t need to be turned off. There are some exceptions though where these preventions impair the operation of certain network designs. eg.. Frame relay network setup in a hub-spoke topology, due to things like split horizon certain routes won’t be advertised to adjacent routers, i’ll explain that later.


EIGRP – Enhanced Interior Gateway Protocol – The Basics

April 1, 2014

EIGRP Notes:

  • Reffered to as a hybrid routing protocol or advanced distance vector routing protocol within some documentation
  • Supports VLSM and route summarisation
  • Cisco proprietary protocol
  • Includes Subnet mask in routing updates

 

EIGRP supports the following features.

  • IPv4 and v6
  • Classless
  • VLSM and CIDR
  • Summarisation of networks and discontiguous ones
  • Uses Reliable Transport Protocol (RTP)
  • Path selection is via Diffusing Update Algorithm (DUAL)
  • No broadcasts

The following three values must match between EIGRP enabled routers before they become neighbours and then they can share routes.

  • Reception of either a Hello or ACK
  • Autonomous System Numbers
  • Same K Values

The Hello timer by default is set to 5 seconds, the Hello message is used to establish neighbour relationships with other routers. The Hellos will alert other routers, also of new routers on the internetwork. This also means that when the Hello messages stop, the neighbour relationship will stop.

This brings me onto the Hold Timer, this timer is used to tell the router the time frame it should wait to receive a Hello from a neighbour before it declares that router dead. Once that happens it will be removed from the neighbour table and any routes which used that now dead router will have their paths recalculated.


Routing Metrics

December 10, 2013

On our journey through the Cisco ICND2 objectives, we are expected to know a thing or two about “Metrics”

So what is a metric you ask ?

Well a routing metric is used by all routing protocols to decide on the best path to any given network. Not to be confused with Administrative Distance (AD). Which tell the router which path to believe when a router learns a route from more then one routing protocol.

Below is an example of what a metric is, with the use of RIPv2. As you can see R1 has two paths to reach R2. One is out FA0/1 directly to R2 (Which in the real world would be a connected route with a lower AD then RIP) and out FA0/2 through R3.

The metric with RIP is calculated using hop count, so the path with the least hops is the best path. The path highlighted in red would be used in this example.

RIP Metric


Split Horizon

December 9, 2013

The next routing protocol feature to talk about is Split Horizon.

A feature of Distance Vector that prevents routing loops from forming. Split horizon rules stop a router from sending routing updates back out the interface from which it learnt them from.

Split horizon is used with RIPv2 and EIGRP and yup you guessed it they are Distance Vector routing protocols.

The example below gives you some understanding of what Split Horizon is actually doing. The RIP update going out R1’s FA0/1 interface only includes the subnet 192.168.0.0/24 (Route highlighted in red). As you can see from the routing table below, R1 learnt this network from an update off FA0/3, so Split Horizon Rules allow this network to be sent in an update out FA0/1.

The routes highlighted in black, are learnt from FA0/1 so Split horizon rules will not allow these routes to be sent in an update, over FA0/1

Split_Horizon


Default Administrative Distances

December 1, 2013

The ICND2 exam asks for knowledge about Routing protocols and their Default Administrative distances. The following table below should be burned into your brain before taking the exam i would imagine.

An Administrative Distance is used so a router can make a decision on how believable the route is as reported by the corresponding routing protocols. Learn this and in the coming days we will use this to configure and Verify both flavours of OSPF and EIGRP.

Route Administrative Distance
Connected 0
Static 1
BGP (External Routes) 20
EIGRP (Internal Routes) 90
IGRP 100
OSPF 110
IS-IS 115
RIP 120
EIGRP (External Routes) 170
BGP (Internal Routes) 200
Unusable 255