Information Technology Reference
In-Depth Information
understanding the beneits of subnetting
To understand one benefit of subnetting, consider a hotel or office building. Say that a
hotel has 1,000 rooms with 75 rooms to a floor. You could start at the first room on the
first floor and number it 1; then when you get to the first room on the second floor, you
could number it 76 and keep going until you reach room 1,000. But someone looking for
room 521 would have to guess on which floor that room is located. If you were to
“subnet” the hotel, you would identify the first room on the first floor with the number
101 (1 = Floor 1 and 01 = Room 1), the first room on the second floor with 201, and so on.
The guest looking for room 521 would go to the fifth floor and look for room 21.
An organization with a single network address (comparable to the hotel building
mentioned in the sidebar “Understanding the Benefits of Subnetting”) can have a subnet
address for each individual physical network (comparable to a floor in the hotel building).
Each subnet is still part of the shared network address, but it also has an additional identifier
denoting its individual subnetwork number. This identifier is called a
subnet address
.
Subnetting solves several addressing problems:
■
If an organization has several physical networks but only one IP network address, it
can handle the situation by creating subnets.
■
Because subnetting allows many physical networks to be grouped together, fewer
entries in a routing table are required, notably reducing network overhead.
■
These things combine collectively to yield greatly enhanced network efficiency.
The original designers of the Internet Protocol envisioned a small Internet with only tens
of networks and hundreds of hosts. Their addressing scheme used a network address for
each physical network. As you can imagine, this scheme and the unforeseen growth of the
Internet created a few problems. The following are two examples:
Not Enough Addresses
A single network address can be used to refer to multiple physical
networks, but an organization can request individual network addresses for each one of its
physical networks. If all of these requests were granted, there wouldn't be enough addresses
to go around.
Gigantic Routing Tables
If each router on the Internet needed to know about every
physical network, routing tables would be impossibly huge. There would be an
overwhelming amount of administrative overhead to maintain those tables, and the
resulting physical overhead on the routers would be massive (CPU cycles, memory, disk
space, and so on). Because routers exchange routing information with each other, an
additional, related consequence is that a terrific overabundance of network traffic would
result.
Search WWH ::
Custom Search