Corporate.Net internetRx By Chris Lewis Q: When we first got connected to the Internet, we had a consultant set up our router and IP addressing scheme. At that time, we had one Class C network address assigned to us by our Internet service provider (ISP), and the consultant used subnet masks to break this one network number into several subnetwork numbers. I know that portions of the network have been split on different segments of the router. What does all that subnetting mean? A: Essentially, your consultant took a Class C and split it into a couple of separate "sub" networks. This means that your 256 addresses have been split into equal groups that can be uniformly used on different, routed networks. Cr eating more than one subnet under a single network number makes traffic more manageable. There are many reasons why you would want to subnet one network number into several subnets. One reason is you use "real" Internet addresses on your network, supplied by an ISP. In this case, if you have 200 PCs to connect to the Internet, you will probably be assigned one Class C from the ISP. To manage broadcast traffic, or to apply different access list security to different groups of PCs, you could segment this one number into several subnets which, with respect to broadcasts, behave as if they were separate networks. In the IP address notation that you are familiar with, you have four distinct numbers that each represent 8 bits of the actual IP address. A Class C address such as the one you have, has the top three numbers fixed. Thus, 24 bits are assigned and cannot be changed. So, you have been assigned a Class C, such as 200 .1.1.0. The last number can be manipulated in any way. The subnet process is most simply observed using numeric examples. So, you have been assigned the Class C network number 200.1.1.0 by the ISP. You can number PCs on your network from 200.1.1.1 up to 200.1.1.254. The 200.1.1.0 value is reserved for the network number (some IP stacks treat this value as a broadcast), and 200.1.1.255 is the broadcast address. Now, let's say that we want to split this one network number into four subnets by applying a subnet mask value to every device interface on this network. We know that from 200.1.1.0 to 200.1.1.255, there are 256 numbers; the zero value being the first number. If we divide 256 by four, we get 64, which tells us that to get four equal-size subnets out of this one network number, we will get subnets numbered as follows: the first subnet ranges from 200.1.1.0 to 200.1.1.63, the second subnet ranges from 200.1.1.64 to 200.1.1.127, the third subnet ranges from 200.1.1.128 to 200.1.1.191, and the four th subnet ranges from 200.1.1.192 to 200.1.1.255. These ranges give us 64 discrete numbers in each range. Just as in the whole network number case, the first and last address in each subnet range must not be assigned to host PCs. Now the big question: What subnet value do we assign to each interface on our network devices t o split up the network this way? The simple way to do it is to subtract the value of discrete numbers you want in each subnet from 256. We want 64 numbers in each range, so 256 minus 64 equals 192. The subnet mask value is 255.255.255.192.

The Dawning of the A ge of Java Management by Bruce Bordman Web Middleware Glue Binds Web Apps by Barry Nance Updated May 12, 1997