Master EIGRP For Your Next Cisco Exam
- select the contributor at the end of the page -
What this means is that it takes attributes from both distance-vector and link-state routing protocol types. If all of your equipment is and will continue to be Cisco then EIGRP is a good option, if your network is not exclusively Cisco then a protocol like OSPF is better as it is a standard and is supported by other vendor's equipment.
EIGRP (and IGRP) is different from OSPF in that it utilizes an autonomous system number which must match between all neighboring routers for routing information to be exchanged.
EIGRP utilizes the Diffusing Update Algorithm (DUAL) to calculate routes and utilizes a metric very similar to that used by IGRP with the exception that it multiplies it by 256 to allow for greater metric variation.
With EIGRP, two different tables are populated with information: the standard neighbor table and a topology table. The topology table contains all of the available routes which are available for EIGRP; when using the command show ip eigrp topology the command will only display those routes which meet the feasibility condition.
Now to make sense out of what feasibility condition is you must understand EIGRP terminology.
|Advertised Distance or Reported Distance||The Advertised Distance (AD) is the distance from a given neighbor to the destination router.|
|Feasible Distance||The Feasible Distance (FD) is the distance from the current router to the destination router|
|Feasibility Condition (Requirement)||Within EIGRP there is requirement which must be met for a route to be considered feasible and loop-free.|
This requirement states that in order for a route to be feasible the Advertised Distance of the alternate route must be lower than that of the Feasible distance of the current route (see example below).
|Successor||The Successor(s) are the current routes which are the best and are entered into the routing table. By default, these are the ones with the lowest metric and which meet the feasibility condition.|
|Feasible Successor||If an alternate route exists and it meets the requirements of the Feasibility Condition then it is considered a Feasible Successor.|
Basic EIGRP (IOS) Configuration
Now to setup the above network the following commands will be required in order for it to work correctly.
router eigrp autonomous-system-id
This command is used to setup the EIGRP process on the router and configure the Autonomous system number used with the network.
network network-id [mask]
The network command is used to setup which networks will be advertised and routed via EIGRP. EIGRP does not require a mask as it will automatically assume a classful boundary; this can be entered in standard form or as an inverse mask and will be displayed in the configuration as an inverse mask.
In this example the no auto-summary command is used, by default EIGRP will create an additional route which will summarize along classful boundaries.
Specific Example Configuration
The specific EIGRP configuration for this example is:
Simple EIGRP metric example
One of the things that seems to be a confusing subject is that of EIGRP metric calculation. When new to it, it is seems as though a math degree is required. However, once a simple run through of the calculation is done then the complexity is shown as nothing but simple individual calculations.
The following is an example of the metric calculations for the sample network:
In this simple network there are two routers which exist and which utilize EIGRP to route between the different networks which they connect to. In this case they are connected via a 1.544 Mbps frame-relay link and each is connected to a separate 100 Mbps Ethernet network.
The Feasible Distances (FD) in this example are correct and are easy to figure out. Each of the Ethernet interfaces has a FD of 28160. As this example is using the default metric calculation only bandwidth and delay are considered.
Metric = 256 x [bandwidth_metric + delay_metric]
bandwidth+metric = 10,000,000 / lowest_path_bandwidth (kbps)
delay_metric = sum_of_path_delay (tens of microseconds)
The bandwidth used is the lowest bandwidth along a path towards a destination, in this case the lowest bandwidth is 100 Mbps (this bandwidth number is taken from the bandwidth configured on the interface and can be changed by using the bandwidth command on the interface).
The bandwidth metric number is calculated by taking 10,000,000 and dividing it by the bandwidth in kbps. So for this example the calculation is 10,000,000 / 100,000 = 100. The delay is calculated by adding up the delay along the entire path and is written in tensof microseconds, this part can be confusing as the delay that is displayed when using IOS is in microseconds and again can be changed on the interface with the delay command. In this case the delay on a 100 Mbps Ethernet interface is 100 microseconds and is written as 10 tens of microseconds. You add the bandwidth and the delay to reach 110 and multiply it by 256 which gives you 28160.
The serial interface FD is calculated the same way. The bandwidth is 1.544 Mbps which is 1,544 kbps which makes the bandwidth metric 10,000,000 / 1544 = 6476 (EIGRP metric calculations always round down). The delay for a 1.544 Mbps frame-relay interface is 20,000 microseconds which is 2,000 tens of microseconds. You add the bandwidth and the delay to reach 8,476 and multiply it by 256 which gives you 2169856.
Now the Advertised Distance (AD) is the metric advertised from a neighboring router. For example, when viewing the EIGRP topology table on R1 the 172.16.1.0 network is being advertised by R2 and has an AD of 28160 (which is R2's FD to the 172.16.1.0 network). From R1's perspective the route for the 172.16.1.0 network meets the feasibility condition because 28160 is less than 2172416 which is R1's FD to the 172.16.1.0 network.