GE GFK-1541B - Manual (Page 180)

GFK-1541B Chapter 6 Network Administration Support 6-5
6
Subnets and Multiple Gateways
For a site requiring a large network (such as a Class A network) the number of entries in the
routing tables of a site’s internal gateways could number in the millions if no internal network
structure is superimposed on the address definition. The solution to this problem is subnetting.
Subnets allow a site’s network administrators to divide a large network into several smaller
networks while still presenting the overall network as one single entity to the outside world. Each
of the site’s interior gateways need only maintain the subnet numbers of other interior gateways
instead of every single host on the entire network.
Subnet Addressing and Subnet Masks
Subnet addressing is an extension of the IP address scheme that allows a site to use a single netid
for multiple physical networks. Routing outside the site continues as usual by dividing the IP
address into a netid and a hostid via the class.
The standard format for the netid bits and hostid bits for an IP address in a Class B network is
shown below.
10000000 00000011 00000000 00000001
netid bits hostid bits
(binary)
Inside a site the subnet mask is used to re-divide the IP address into a custom netid portion and
hostid portion. Consider adding another physical network to Network 2 (a Class B network) in
the previous example (see Figure 7-2). The result is shown in Figure 7-3. Selecting the subnet
mask shown below would add two additional netid bits allowing for four physical networks
addressed as 0, 64, 128, and 192. The added subnet bits are normally taken from the hostid bits
adjacent to the netid and the subnet mask identifies these bits.
11111111 11111111 11000000 00000000 = 255.255.192.0
netid bits hostid bits subnet mask
(dotted decimal) (binary)
The bits in the subnet mask correspond one to one with the Internet address. The bits in the mask
that are 1 treat the corresponding bits in the IP address as part of the netid bits. The bits in the
mask that are 0 treat the corresponding bits as part of the hostid bits.
In effect, two bits of the Class B hostid have been used to extend the netid, creating an extended
netid
, or subnetid. Each unique combination of bits in the part of the hostid where subnet mask
bits are 1 specifies a different physical network.