Illustrated TCP/IP by Matthew G. Naugle Wiley Computer Publishing, John Wiley & Sons, Inc. ISBN: 0471196568   Pub Date: 11/01/98

Chapter 93
Direct Routing

As stated before there are two type of routing: direct and indirect. This section gives you a brief introduction to direct routing. Throughout this section, different network numbers will be used. The examples will not employ the use of subnets. Subnets effectively act like network numbers. Subnets are also separated by a router. For example, in the slide, the network numbers could be 140.1.1.1 on the network with endstation B, and 140.1.2.1 on the network containing host A. Using a subnet mask of 255.255.255.0 would yield two different networks: 140.1.1.0 and 140.1.2.0. For simplicity in explaining routers, I have chosen to use completely different network numbers.

How does a network station know whether the packet has to be directly (local) or indirectly (remote) routed? For the network station, it is a relatively simple process. The whole basis for routing is in the IP network number assigned to the network station.

Remember from the previous section on Addressing that an IP address contains the network number as well as the host number. With the first 1, 2, 3, or 4 bits of the 32-bit IP network address identifying the class of the address, this allows for any network station (workstation or router) to quickly extract the network portion out of the class of IP address. In other words, by reading up to the first 4 bits of the IP address, a network station can quickly determine how much of the IP address to read to determine the network number of the address. The sending station will compare the packet’s destination network number to that of its own network number. If the network number portion of the destination IP address matches it own, the packet can be routed directly on the local LAN, without the use of a router. The packet is simply transmitted to the station (using ARP if necessary).

Once this determination is made, and the packet is destined for a local route, the network station would check its ARP table to find the IP-to-physical-address mapping. If one is found, the packet is physically addressed and transmitted onto the network. The physical destination address (located in the datalink header) will be that of the receiving station. If the station’s address is not in the ARP cache, the ARP request process is invoked.

Referring to the slide, endstation B and host A are located on the same network.

Again, a point needs to be brought up here: There is a difference between a routing protocol and a routable protocol. A routable protocol is one that allows for routing such as NetWare (IPX) and TCP/IP. NetBIOS and LAT (a DEC terminal/printer protocol) are not routable protocols. RIP and OSPF, are routing protocols whichenable the routing functions to work properly.

Direct Routing