posted 22 years ago
The subnet mask is a number that helps split your IP address into two parts: a network number and a station number.
In conventional IPv4 usage, the number itself specified a class for this split, based on the first (high-order) three bits of a 32-bit IP number. If the first bit is zero, you have a class A address, meaning the high 8 bits of the IP are the network number, e.g.:
That's a class A address. Class B addresses reserve the high 16 bits to describe the network number, the lower 16 for a station. Finally, class C addresses use the high 24 bits for the network, the last 8 for the station.
In modern small networks, subsets other than these canoncial classes get used so that you can get more than one "subnetwork" out of a class address.
So if your system doesn't agree with the others on your subnet which part of the number is network, and which part of the number defines a station, you aren't going to be able to hear them because you're not going to recognize them.
I don't know how Linux does this. In Solaris, we use ifconfig(1M) to change an adapter dynamically, and we define subnets in a file called /etc/netmasks. I'd look for those two things first, and failing that, ask somebody who knows what the Linux equivalents for this program and file are called.
Make visible what, without you, might perhaps never have been seen.
- Robert Bresson