Each gateway has a defined set of routing tables that tells it the path to certain destinations. Gateways don’t know where every IP address is, so gateways in turn have their own gateways. When a gateway doesn’t know the way to a particular destination, it forwards the datagram to its own gateway. This cycle of routing is continued until the datagram finally arrives at its destination. The entire path traveled from the origin to the destination is called the route.
Datagrams may travel to a particular destination one way one time and another way another time. The route is determined by many variables that occur along the route. For instance, overloaded gateways processing too much traffic may not respond in a timely manner or may simply refuse to route your traffic, causing a timeout. That timeout causes the sending gateway to seek an alternative route for the datagram. A timeout could also be caused by network links or network devices being down or many other factors.
In most cases, routes can be predefined and made to remain static if need be. Alternative routes can also be predefined, providing a maximum probability that your datagrams will arrive at their final destination. The way routing is handled on your network is purely a matter of preference. You should note that improper routing can lead to poor network performance; network routers may become overburdened with traffic or may inadvertently assign longer routes than necessary.
Hosts, Host Names, and Domain Names
The word host describes various devices on the network. Technically, any device on a network that has some sort of address and is capable of sending and/or receiving data or information can be called a host. A printer can be a host, a network fax machine can be a host, and a computer system can certainly be a host. However, hosts are usually thought of as computer systems on a network. On most networks, hosts have names. On TCP/IP networks, hosts have names that are a subset of domain names.
0 comments:
Post a Comment