the network.
The processors in each node handle this task, which takes away from the processing power needed for other
tasks and application—thus causing a slowdown that the users discover and complain about. Most network
administrators pass off this slowness as a problem with the PCs, and the most vital PCs are rebuilt or replaced.
When the companies finally decide to upgrade to a switched network, they can typically do so over a
weekend. When the network users leave on Friday, their high−powered Pentiums stacked with RAM have the
speed of 386s. When they return Monday morning, nothing is more exciting than hearing comments all over
the office about how their computers boot up more quickly and run so much faster, and how they like the
faster network. But did the users get a faster network? In one sense, the network did get an upgrade; but this
upgrade merely eliminated the problems of a flat topology network by segmenting the network into smaller
collision and broadcast domains.
How did they do this? By replacing the hubs (which send data they receive out every single port, forcing
every node attached to them to process the data whether the node is meant to receive the data or not) with
switches. In terms of per−port costs, replacing your hubs with switches is a solution at a quarter of the cost of
upgrading the network cabling. So, what segments the network? VLANs.
Note Sometimes, if you have a 10BaseT network with Category 3 or 4 cabling, the best solution is
to fix the immediate problems by upgrading to Category 5 cabling and implementing a Fast
Ethernet network in conjunction with installing switches. However, most network users do
not need more than true 10Mbps from the Access layer switches to their desktops even if
they are using high−bandwidth applications. After all, before they had switches, the users
were getting along with only 3Mbps or 4Mbps on their 10Mbps link, due to broadcasts,
collisions, and network utilization.
Why Use VLANs?
VLANs are used to segment the network into smaller broadcast domains or segments. The primary reason to
segment your network is to relieve network congestion and increase bandwidth. Segmentation is often
necessary to satisfy the bandwidth requirements of a new application or a type of information the network
needs to be able to support, such as multimedia or graphical design applications. Other times, you may need to
segment the network due to the increased traffic on the segment or subnet.
Be careful not to oversegment. Placing each port in an individual VLAN is like placing a router to stop
broadcasts between each individual VLAN. Routers are like bug poison—they kill broadcasts dead.
Broadcasts can’t escape through routers and they can’t escape a VLAN, either. Each VLAN becomes its own
individual broadcast domain. When a network node or workstation sends out an advertisement or broadcast to
the other nodes on a segment, only the nodes assigned to the VLAN to which the node sending the broadcast
is assigned will receive that broadcast.
Another definition of a VLAN is a logical grouping of network users and resources connected
administratively to defined ports on a switch. By creating VLANs, you are able to create smaller broadcast
domains within a switch by assigning different ports on the switch to different subnetworks. Ports assigned to
a VLAN are treated like their own subnet or broadcast domain. As a result, frames broadcast are only
switched between ports in the same VLAN at Layer 2.
Using virtual LANs, you’re no longer confined to physical locations. VLANs can be organized by location,
function, department, or even the application or protocol used, regardless of where the resources or users are
located. In a flat network topology, your broadcast domain consists of all the interfaces in your segment or
subnet. If no devices—such as switches or routers—divide your network, you have only one broadcast
domain. On some switches, an almost limitless number of broadcast domains or VLANs can be configured.
89