NetBEUI is an example of an non-routable protocol. NetBEUI is short for NetBIOS Enhanced User Interface. It ships with all Microsoft Windows operating systems. In small, peer-to-peer networks, NetBEUI is easy to configure and use.
A routed protocol is used between routers to direct user traffic. Examples of routed protocols are: IP, AppleTalk, and IPX. These protocols can be routed at the same time by the same router. We say, routers are capable of supporting multiple independent routing protocols and maintaining routing tables for several routed protocols concurrently.
Routing protocols work between routers to maintain path tables. Path determination occurs at Layer 3, the network layer. Routing tables store information on possible destinations and how to reach each of the destinations. Routing tables are used to forward packets from the source network to the destination network.
Consistency of Layer 3 addresses across the entire internetwork also improves the use of bandwidth by preventing unnecessary broadcasts. Broadcasts invoke unnecessary process overhead and waste capacity on any devices or links that do not need to receive the broadcast.
As you know, most network protocol addressing schemes are broken down into two parts: Network and host.
To relay a packet, a router uses two basic functions: a path determination function and a switching function. The switching function allows a router to accept a packet on one interface and forward it on a second interface.
Dynamic routing works differently. Route knowledge is updated automatically for topology changes by a routing process. Dynamic routing depends on two basic router functions: maintaining a routing table and timely updates.
Default routes are routes defined by the system administrator as the path to take when no route to the destination is known.
Distance-vector protocols periodically broadcast entire routing tables to neighbor routers. Distance vector routing begins by identifying directly connected networks and topology changes proceed in distance-vector step by step from router to router.
Link-state protocols broadcast updates to ALL other routers on the internetwork upon startup and when the network topology changes. Link-state is also known as the shortest path first routing algorithm.
A link-state algorithm allows a router to know the exact topology of the entire internetwork.
Distance-vector algorithms do not allow a router to know the exact topology of an internetwork.
Major characteristics of distance-vector and link-state routing protocols:
Distance-vector
- Periodically broadcast entire routing table to neighbors
- Slow to converge
- Prone to routing loops because of their routing by rumor nature
- Easy to configure and administer
Link-State
- Broadcast entire routing table to all other routers in the autonomous system (AS) on start up, all other routing table updates contain only updated routes, updates only occur when a network topology change occurs
- Fast to converge due to link-state advertisements
- Less prone to routing loops because all other routers share a common view of the network
- Harder to configure; requires greater memory and processing power on each router
Hop count is defined as the number of routers a packet must pass through.
RIP - a distance vector routing protocol. RIP updates occur every 30 seconds.
IGRP - Cisco's distance vector routing protocol. Routing updates occur every 90 seconds.
OSPF - A link-state routing protocol.
EIGRP - A balanced hybrid routing protocol.
IGRP is also a distance-vector routing protocol. IGRP has a maximum hop count of 255. IGRP is not limited to using hop count as its sole metric. IGRP can also use load, bandwidth, reliability, delay, and maximum transmission unit when determining best path.
Administrative Distances
Route Learned Via:
Administrative Distance
Direction Connection 0
Static route
1
IGRP
100
OSPF
110
RIP
120
Administrative distance is a value used to determine the reliability of a particular route. The above information shows common routing protocols and their administrative distances. Therefore, if a route is being discovered using both RIP and a directly connected interface, the route available via the directly connected interface will be the preferred route because it has a lower administrative distance. Likewise, if both IGRP and RIP advertise a route for a particular network, the IGRP route will be used because it is considered more reliable due to its lower administrative distance.
Each routing protocol interprets best path in its own way. A variety of metrics can be used to define the best path. Some routing protocols, such as RIP, use only one metric, and some routing protocols, such as IGRP, use a combination of metrics.
Commonly Used Metrics
Hop count - The number of routers a packet must go through to reach a destination. The lower the hop count the better the path.
Bandwidth - The data capacity of a link.
Delay - The length of time required to move a packet from source to destination.
Load - The amount of activity on a network resource, such as a router or link.
Reliability - The error rate of each network link.
Ticks - The delay on a data link using IBM PC clock ticks (approximately 55 milliseconds).
Cost - The arbitrary value, usually based on bandwidth, dollar expense, or other measurement, that is assigned by a network administrator.
Routing loops are often referred to as count-to-infinity problems because loops, without preventive measures, will cause packets to bounce around the internetwork for an infinite amount of time. To prevent these problems, techniques such as defining a maximum, split horizon, split horizon with poison reverse, and hold-down timers are used to reduce the chances that incorrect route table information will be propagated.
Defining a maximum is one of the easiest ways to limit count-to-infinity problems If you assign a packet a maximum hop count, it cannot bounce infinitely around the internetwork. RIP, one of the most common distance-vector protocols, defines a maximum hop count of 15. Therefore, if a routing loop did occur on a RIP internetwork, the packet would only travel through 15 hops before the packet exceeded its time to live (TTL) and dropped. In other words, the 16th router that the packet tried to cross on a RIP internetwork would see that the packet had exceeded its TTL and would drop the packet. Time to live (TTL) is normally the same as the hop count. A packet with a TTL of 15 can pass through 15 routers before it is dropped.
Split horizon and split horizon with poison reverse are two other common ways to prevent routing loops. Split horizon controls what information a router will send out about particular routes. In short, routers will not send information back through an interface about an advertised route from that interface. For example, a router will not send an update for a route via an interface from which it has received knowledge of that route.
If a router uses split horizon with poison reverse it not only refuses to send information back through an interface that advertised the route but also responds to the router that attempted to update it. In other words, it poisons the erroneous route advertised by that one router so that no other router will see this as a viable route.
Another common technique used to stop routing loops is the hold-down timer. Hold-down timers allow a router to place a route in a state where it will not accept any changes to a route. Updates are ignored because the route would in "hold down" for a period of time after it was marked down.
Routers configured with a link-state routing protocol use Link-state advertisements (LSAs) to inform all routers on the internetwork of their route tables. Link-state packets (LSPs), packets used to send out LSAs, allow every router in the internetwork to share a common view of the topology of the internetwork. A router configured with a link-state routing protocol floods or broadcasts LSPs to the network so that every other router on the internetwork has a common view of the topology of the internetwork. The network quickly reaches a state of convergence due to the flooding of link-state packets. This is one huge advantage link-state routing protocols have over distance-vector routing protocols. Also, later updates by the routers in the internetwork will be triggered updates. These updates occur due to network topology changes, not periodic route table advertisements. This announcement contains only the changes in the route table, not the entire route table. Link-state routing protocols do have some problems. Due to the complexity of the Shortest Path First algorithm and the need to hold a view of the entire internetwork topology, routers using link-state protocols most be configured with more memory and processing power than those using distance-vector routing protocols.