116
6.2 Subnet masks
A mask looks like a regular IP address, but contains a
pattern of bits that tells what parts of an IP address are the
network ID and what parts are the host ID: bits set to 1
mean "this bit is part of the network ID" and bits set to 0
mean "this bit is part of the host ID."
Subnet masks are used to define subnets (what you get after dividing a
network into smaller pieces). A subnet's network ID is created by "borrowing"
one or more bits from the host ID portion of the address. The subnet mask
identifies these host ID bits.
For example, consider a class C network 192.168.1. To split this into two
subnets, you would use the subnet mask:
255.255.255.128
It's easier to see what's happening if we write this in binary:
11111111. 11111111. 11111111.10000000
As with any class C address, all of the bits in field1 through field 3 are part of
the network ID, but note how the mask specifies that the first bit in field 4 is also
included. Since this extra bit has only two values (0 and 1), this means there
are two subnets. Each subnet uses the remaining 7 bits in field4 for its host IDs,
which range from 0 to 127 (instead of the usual 0 to 255 for a class C address).
Similarly, to split a class C network into four subnets, the mask is:
255.255.255.192 or 11111111. 11111111. 11111111.11000000