In 1984, Cisco created the Interior Gateway Routing Protocol (IGRP) to address issues with RIP in large networks. IGRP is a distance vector protocol, however, it employs several routing metrics (not just hop count) to compute the destination's distance. Hold-downs, split horizons, and poison-reverse updates are IGRP features aimed at improving network stability. IGRP should only be utilized if your current environment consists solely of IGRP and you do not wish to add another routing protocol.

The IGRP protocol offers the following routing goals:

• The capacity to manage many "types of services" with a single set of data
• 
  
• Routing loop prevention
• 
  
• Routing stability, even in extremely large or complex networks
• 
  
• Low overhead, indicating that IGRP should not consume more bandwidth than it needs for its own operation
• 
  
• Rapid reaction to varying network structure
• 
  
• Split traffic along parallel routes when their desirability is equal.
• 
  
• Consideration of error rates and traffic levels on various paths

IGRP is a distance-vector protocol in which routers (commonly referred to as gateways) only exchange routing information with neighboring routers. IGRP outperforms RIP in terms of metrics. It utilizes many of RIP's fundamental functionalities but increases the maximum number of hops supported to 100. Consequently, it may function better on larger networks. IGRP compares network parameters such as capacity, dependability, and load to function. This type automatically updates when modifications, such as route modifications, occur. This aids in the prevention of routing loops, which are faults that result in an infinite loop of data transfer. The new IGRP measures include the following:

• Bandwidth of the path section with the smallest bandwidth. The transmission rate in bits per second.
• 
  
• Topological delay time. The time it would take for a packet to reach its destination if the network were not crowded. If there is network traffic on the network, you may experience additional delays.
• 
  
• Dependability of the route. Indicates the path's reliability based on the number of packets that really arrive at the destination, relative to the total number of packets transmitted.
• 
  
• Path occupancy of the channel. Indicates the percentage of bandwidth currently in use. This value will fluctuate frequently as network traffic fluctuates.

Using a complex algorithm, IGRP evaluates these parameters and determines the optimal route, as represented by the smallest metric value.

Hold-downs, Split horizons, and Poison-reverse updates are further significant stability characteristics of IGRP.

• Hold-downs: Used to prevent a regular update message from reestablishing a route that may have previously become invalid. When a network link fails, surrounding routers will detect the absence of regularly scheduled updates and determine that the link is no longer operational. The network will subsequently begin to propagate messages informing users that this router is not operating. If this convergence takes too long, another router on the network may indicate that this router is still operating normally. This gadget may be broadcasting inaccurate route information. A hold-down instructs the network's routers to delay for a period of time any modifications that could disrupt the routes. The hold-down duration is calculated to be only marginally longer than the time required to update the entire network with a route change.
• 
  
• Split horizons: Used to prevent routing loops from occurring between two routers. It is never advantageous to relay route information back in the direction from whence a packet was sent.
• 
  
• Poison-reverse updates: Used to reduce loops between many routers. When the metric rises dramatically, it may suggest a routing loop. The router is subsequently placed on hold-down by sending it a poison-reverse update.

Utilizing timers and variables containing time intervals is another characteristic of IGRP's stability. Included among the timers are as follows:

• Update Timer: The update timer specifies how frequently update messages are transmitted. The IGRP default update interval is 90 seconds.
• 
  
• Invalid Timer: The invalid timer specifies how long a router will wait before declaring a route invalid if it is not receiving routing update messages. The default value for the IGRP invalid timer is three times the update timer.
• 
  
• Hold-time Period: The hold-time period (also known as the hold-down period) will indicate the duration of the hold-down period. The default hold-time for IGRP is three times the update interval plus ten seconds.
• 
  
• Flush Timer: The flush timer specifies the amount of time that must elapse before a route is removed from a routing database. The default value of the IGRP flush timer is seven times the update interval.
• 
  
• Sleep Timer: The sleep timer specifies how long update messages will be delayed. The sleep value should be less than the update timer; otherwise, the routing tables will never be synchronized.