I'm trying to get a grasp on IP addressing concepts. What I understand is that there used to be classes of IP addresses: A, B, and C. All IP addresses had a network identifier and then a host. Then they invented subnetting. An IP address was a network, a subnet, and then a host. So The Internets could route to the network, and the network would route to the subnet, which would then route to the host.
So if I have a subnet I have to have a router for it. And I also have to have a subnet mask. The subnet mask is either 255.255.255.0 (or 255.255.0.0, etc) where the 0's indicate the portion for host identification, or the subnet mask can be written at the network name with a slash and then a number indicating the length of the network identifier. For example: 63.98.246.161/8? That doesn't make sense to me. How could I specify a whole address but then also have room for a mask? Then there's the concept of variable length subnet masks, which I don't have to worry about. Classless routing is called "CIDR". I'm in an environment with 67 subnets and I'm configuring a device scanner for this environment. I'm defining one configuration file for each subnet so that when I scan it will be one subnet at a time, instead of just push and go and tear up the network. I've been given a list of subnet addresses and their masks. Most of them are of the type: 167.190.45.0 (address) and 255.255.255.0 (mask) which yields an IP range of 167.190.45.1 up to 167.190.45.254. That's fine. But then some of them are of the type: 167.190.1.24 (address) and 255.255.255.248 (mask) which yields an IP range of... ? Sometimes the mask is like 255.255.240.0 or 255.255.255.192. Aagh! What the heck is going on? Is there a calculator for this stuff? I would really appreciate it if one of you network brains could explain this odd subnet masking or explain where the kink is in my understanding. -todd -- KPLUG-List@kernel-panic.org http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
