Routing protocol algorithms use metrics, which are numerical values that are associated with specific routes. These values are used to prioritize of prefer routes learned by the routing protocol from the most preferred to the least preferred. In essence, the lower the route metric, the more preferred the route by the routing protocol. The route with the lowest metric is typically the route is the least cost or best route to the destination network. This route will be placed into the routing table and be used to forward packets to the destination network.
Different routing algorithms use different variables to compute the route metric. Some routing algorithms use only a single variable, while other advanced routing protocols may use more than one variable to determine the metric for a particular route. In most cases, the metrics that are computed by one routing protocol are incompatible with those used by other routing protocols. The different routing protocol metrics may be based on one or more of the following:
- Bandwidth
- Cost
- Delay
- Load
- Path Length
- Reliability
Bandwidth
The term bandwidth refers to the amount of data that can be carried from one point to another in a given time period. Routing algorithms may use bandwidth to determine which link type is preferred over another. For example, a routing algorithm might prefer a GigabitEthernet link over a FastEthernet link because of the increased capacity of the GigabitEthernet link over the FastEthernet link.
In Cisco IOS software, the bandwidth interface configuration command can be used to adjust the default bandwidth value for an interface, effectively manipulating the selection of one interface against another by a routing algorithm. For example, if the FastEthernet interface was configured with the bandwidth 1000000interface configuration command, both the FastEthernet and GigabitEthernet links would appear to have the same capacity to the routing algorithm and would be assigned the same metric value. The fact that one of the links is actually a FastEthernet interface while another is actually a GigabitEthernet link is irrelevant to the routing protocol.
From a network administrator’s point-of-view, it is important to understand that the bandwidth command does not affect the physical capability of the interface. In other words, configuring the higher bandwidth on the FastEthernet interface does not mean that it is capable of supporting GigabitEthernet speeds. OSPF and EIGRP use bandwidth in metric calculation.
Cost
The cost, as it pertains to routing algorithms, actually refers to communication cost. This refers to the actual monetary cost of link utilization. The cost may be used when, for example, a company prefers to route across private links rather than public links that include monetary charges for sending data across them or for the usage time. Intermediate System-to- Intermediate System (IS-IS) supports an optional expense metric measures the monetary cost of link utilization. The IS-IS protocol is beyond the scope of the ROUTE exam and will not be described in any detail in this guide.
Delay
There are many types of delay, all of which affect different types of traffic. In general, delay refers to the length of time required to move a packet from source to destination through the internetwork. In Cisco IOS software, the interface delay value is in microseconds.
The interface value is configured using the delay interface configuration command. When you configure the interface delay value, it is important to remember that this does not affect traffic. For example, configuring a delay value of 5000 does not mean that traffic sent out of that interface will have an additional delay of 5000. Table 1-2 shows the default delay values for common interfaces in Cisco IOS software:
| Interface Type | Delay (usec) |
| 10 Mbps Ethernet | 1000 |
| FastEthernet | 100 |
| GigabitEthernet | 10 |
| T1 Serial | 20000 |
Tab.1-2. Interface Delay Values
EIGRP uses the interface delay value as part of its metric calculation. Manually adjusting the interface delay value results in the re-computation of the EIGRP metric.
Load
The term load means different things to different people. For example, in general computing terminology, the term load refers to the amount of work a resource, such as the CPU, is performing. The term, as it applies in this context, refers to the degree of use for a particular router interface. The load on the interface is as a fraction of 255. For example, a load of 255/255 indicates that the interface is completely saturated, while a load of 128/255 indicates that the interface is 50% saturated. By default, the load is calculated as an average over a period of five minutes. The interface load value can be used by EIGRP in its metric calculation.
Path Length
The path length metric is the total length of the path that is traversed from the local router to the destination network. Different routing algorithms represent this in different forms. For example, Routing Information Protocol (RIP) counts all intermediate routers (hops) between the local router and the destination network and uses the hop count as the metric, while Border Gateway Protocol (BGP) counts the number of traversed autonomous systems between the local router and destination network and uses the autonomous system count to select the best path.
Reliability
Like load, the term reliability means different things depending on the context in which it is used. In this guide, unless stated otherwise, it should always be assumed that the term reliability refers to the dependability of network links or interfaces. In Cisco IOS software, the reliability of a link or interface is represented as a fraction of 255. An as example, a reliability value of 255/255 indicates the interface is 100% reliability. Similar to the interface load, by default the reliability of an interface is calculated as an average over a period of five minutes.