• File Servers

• Workstations

• Network Interface Cards

• Concentrators/Hubs

• Repeaters

• Bridges

• Routers

File Servers
A file server stands at the heart of most networks. It is a very fast computer with a large
amount of RAM and storage space, along with a fast network interface card. The network
operating system software resides on this computer, along with any software applications
and data files that need to be shared.

The file server controls the communication of information between the nodes on a
network. For example, it may be asked to send a word processor program to one
workstation, receive a database file from another workstation, and store an e-mail
message during the same time period. This requires a computer that can store a lot of
information and share it very quickly. File servers should have at least the following
characteristics:

• 75 megahertz or faster microprocessor (Pentium, PowerPC)

• A fast hard drive with at least four gigabytes of storage
• A RAID (Redundant Array of Inexpensive Disks) to preserve data after a disk
casualty
• A tape back-up unit
• Numerous expansion slots
 • Fast network interface card
• At least of 32 MB of RAM

Workstations
All of the computers connected to the file server on a network are called workstations. A
typical workstation is a computer that is configured with a network interface card,
networking software, and the appropriate cables. Workstations do not necessarily need
floppy disk drives or hard drives because files can be saved on the file server. Almost any
computer can serve as a network workstation.

Network Interface Cards

The network interface card (NIC) provides the physical connection between the network
and the computer workstation. Most NICs are internal, with the card fitting into an
expansion slot inside the computer. Some computers, such as Mac Classics, use external
boxes which are attached to a serial port or a SCSI port. Laptop computers generally use
external LAN adapters connected to the parallel port or network cards that slip into a
PCMCIA slot.

Network interface cards are a major factor in determining the speed and performance of a
network. It is a good idea to use the fastest network card available for the type of
workstation you are using.

The three most common network interface connections are Ethernet cards, LocalTalk
connectors, and Token Ring cards. According to a International Data Corporation study,
Ethernet is the most popular, followed by Token Ring and LocalTalk (Sant'Angelo, R.
(1995). NetWare Unleashed, Indianapolis, IN: Sams Publishing).

Ethernet Cards

Ethernet cards are usually purchased separately from a computer, although many
computers (such as the Macintosh) now include an option for a pre-installed Ethernet
card. Ethernet cards contain connections for either coaxial or twisted pair cables (or both)
(See fig. 1). If it is designed for coaxial cable, the connection will be BNC. If it is
designed for twisted pair, it will have a RJ-45 connection. Some Ethernet cards also
contain an AUI connector. This can be used to attach coaxial, twisted pair, or fiber optics
cable to an Ethernet card. When this method is used there is always an external
transceiver attached to the workstation. (See the Cabling section for more information on
connectors.)
 Fig. 1. Ethernet card.
From top to bottom:
RJ-45, AUI, and BNC connectors

LocalTalk Connectors

LocalTalk is Apple's built-in solution for networking Macintosh computers. It utilizes a

special adapter box and a cable that plugs into the printer port of a Macintosh (See fig. 2).
A major disadvantage of LocalTalk is that it is slow in comparison to Ethernet. Most
Ethernet connections operate at 10 Mbps (Megabits per second). In contrast, LocalTalk
operates at only 230 Kbps (or .23 Mbps).

Fig.2. LocalTalk connectors

Ethernet Cards vs. LocalTalk Connections

Ethernet LocalTalk
Fast data transfer (10 Mbps) Slow data transfer (.23 Mbps)
Expensive - purchased separately Built into Macintosh computers
Requires computer slot No computer slot necessary
Available for most computers Works only on Macintosh computers
Token Ring Cards

Token Ring network cards look similar to Ethernet cards. One visible difference is the
type of connector on the back end of the card. Token Ring cards generally have a nine pin
DIN type connector to attach the card to the network cable.

Concentrators/Hubs
A concentrator is a device that provides a central connection point for cables from
workstations, servers, and peripherals. In a star topology, twisted-pair wire is run from
each workstation to a central concentrator. Hubs are multislot concentrators into which
can be plugged a number of multi-port cards to provide additional access as the network
grows in size. Some concentrators are passive, that is they allow the signal to pass from
one computer to another without any change. Most concentrators are active, that is they
electrically amplify the signal as it moves from one device to another. Active
concentrators are used like repeaters to extend the length of a network. Concentrators are:

• Usually configured with 8, 12, or 24 RJ-45 ports

• Often used in a star or star-wired ring topology
• Sold with specialized software for port management
• Also called hubs
• Usually installed in a standardized metal rack that also may store netmodems,
bridges, or routers

Repeaters
When a signal travels along a cable, it tends to lose strength. A repeater is a device that
boosts a network's signal as it passes through. The repeater does this by electrically
amplifying the signal it receives and rebroadcasting it. Repeaters can be separate devices
or they can be incorporated into a concentrator. They are used when the total length of
your network cable exceeds the standards set for the type of cable being used.

A good example of the use of repeaters would be in a local area network using a star
topology with unshielded twisted-pair cabling. The length limit for unshielded twisted-
pair cable is 100 meters. The most common configuration is for each workstation to be
connected by twisted-pair cable to a multi-port active concentrator. The concentrator
regenerates all the signals that pass through it allowing for the total length of cable on the
network to exceed the 100 meter limit.

Bridges

Routers
A router translates information from one network to another; it is similar to a
superintelligent bridge. Routers select the best path to route a message, based on the
destination address and origin. The router can direct traffic to prevent head-on collisions,
and is smart enough to know when to direct traffic along back roads and shortcuts.

While bridges know the addresses of all computers on each side of the network, routers
know the addresses of computers, bridges, and other routers on the network. Routers can
even "listen" to the entire network to determine which sections are busiest -- they can
then redirect data around those sections until they clear up.

If you have a school LAN that you want to connect to the Internet, you will need to
purchase a router. In this case, the router serves as the translator between the information
on your LAN and the Internet. It also determines the best route to send the data over the
Internet. Routers can:

• Direct signal traffic efficiently

• Route messages between any two protocols
• Route messages between linear bus, star, and star-wired ring topologies
• Route messages across fiber optic, coaxial, and twisted-pair cabling

Unshielded Twisted Pair (UTP) Cable

Twisted pair cabling comes in two varieties: shielded and unshielded. Unshielded twisted
pair (UTP) is the most popular and is generally the best option for school networks (See
fig. 1).

Fig.1. Unshielded twisted pair

The quality of UTP may vary from telephone-grade wire to extremely high-speed cable.
The cable has four pairs of wires inside the jacket. Each pair is twisted with a different
number of twists per inch to help eliminate interference from adjacent pairs and other
electrical devices. The EIA/TIA (Electronic Industry Association/Telecommunication
Industry Association) has established standards of UTP and rated five categories of wire.

Categories of Unshielded Twisted Pair

Type Use
 Category 1 Voice Only (Telephone Wire)
Category 2 Data to 4 Mbps (LocalTalk)
Category 3 Data to 10 Mbps (Ethernet)
Category 4 Data to 20 Mbps (16 Mbps Token Ring)
Category 5 Data to 100 Mbps (Fast Ethernet)

One difference between the different categories of UTP is the tightness of the twisting of
the copper pairs. The tighter the twisting, the higher the supported transmission rate and
the greater the cost per foot. Buy the best cable you can afford; most schools purchase
Category 3 or Category 5. Category 5 cable is highly recommended.

If you are designing a 10 Mbps Ethernet network and are considering the cost savings of
buying Category 3 wire instead of Category 5, remember that the Category 5 cable will
provide more "room to grow" as transmission technologies increase. Both category 3 and
category 5 UTP have a maximum segment length of 100 meters. In Florida, Category 5
cable is required for retrofit grants. 10BaseT refers to the specifications for unshielded
twisted pair cable (category 3, 4, or 5) carrying Ethernet signals.

Unshielded Twisted Pair Connector

The standard connector for unshielded twisted pair cabling is an RJ-45 connector. This is
a plastic connector that looks like a large telephone-style connector (See fig. 2). A slot
allows the RJ-45 to be inserted only one way. RJ stands for Registered Jack, implying
that the connector follows a standard borrowed from the telephone industry. This
standard designates which wire goes with each pin inside the connector.

Fig.2. RJ-45 connector

Shielded Twisted Pair (STP) Cable

A disadvantage of UTP is that it may be susceptible to radio and electrical frequency
interference. Shielded twisted pair (STP) is suitable for environments with electrical
interference; however, the extra shielding can make the cables quite bulky. Shielded
twisted pair is often used on networks using Token Ring topology.
Coaxial Cable
Coaxial cabling has a single copper conductor at its center. A plastic layer provides
insulation between the center conductor and a braided metal shield (See fig. 3). The metal
shield helps to block any outside interference from fluorescent lights, motors, and other
computers.

Fig.3. Coaxial cable

Although coaxial cabling is difficult to install, it is highly resistant to signal interference.

In addition, it can support greater cable lengths between network devices than twisted
pair cable. The two types of coaxial cabling are: thick coaxial and thin coaxial.

Thin coaxial cable is also referred to as thinnet. 10Base2 refers to the specifications for
thin coaxial cable carrying Ethernet signals. The 2 refers to the approximate maximum
segment length being 200 meters. In actual fact the maximum segment length is 185
meters. Thin coaxial cable is popular in school networks, especially linear bus networks.

Thick coaxial cable is also referred to as thicknet. 10Base5 refers to the specifications for
thick coaxial cable carrying Ethernet signals. The 5 refers to the maximum segment
length being 500 meters. Thick coaxial cable has an extra protective plastic cover that
helps keep moisture away from the center conductor. This makes thick coaxial a great
choice when running longer lengths in a linear bus network. One disadvantage of thick
coaxial is that it does not bend easily and is difficult to install.

Coaxial Cable Connectors

The most common type of connector used with coaxial cables is the Bayone-Neill-
Concelman (BNC) connector (See fig. 4). Different types of adapters are available for
BNC connectors, including a T-connector, barrel connector, and terminator. Connectors
on the cable are the weakest points in any network. To help avoid problems with your
network, always use the BNC connectors that crimp, rather than screw, onto the cable.

Fig.4. BNC connector

Fiber Optic Cable
Fiber optic cabling consists of a center glass core surrounded by several layers of
protective materials (See fig. 5). It transmits light rather than electronic signals,
eliminating the problem of electrical interference. This makes it ideal for certain
environments that contain a large amount of electrical interference. It has also made it the
standard for connecting networks between buildings, due to its immunity to the effects of
moisture and lighting.

Fiber optic cable has the ability to transmit signals over much longer distances than
coaxial and twisted pair. It also has the capability to carry information at vastly greater
speeds. This capacity broadens communication possibilities to include services such as
video conferencing and interactive services. The cost of fiber optic cabling is comparable
to copper cabling; however, it is more difficult to install and modify. 10BaseF refers to
the specifications for fiber optic cable carrying Ethernet signals.

Fig.5. Fiber optic cable

Facts about fiber optic cables:

• Outer insulating jacket is made of Teflon or PVC.

• Kevlar fiber helps to strengthen the cable and prevent breakage.
• A plastic coating is used to cushion the fiber center.
• Center (core) is made of glass or plastic fibers.

Fiber Optic Connector

The most common connector used with fiber optic cable is an ST connector. It is barrel
shaped, similar to a BNC connector. A newer connector, the SC, is becoming more
popular. It has a squared face and is easier to connect in a confined space.

Ethernet Cable Summary

Specification Cable Type Maximum length

10BaseT Unshielded Twisted Pair 100 meters
10Base2 Thin Coaxial 185 meters
10Base5 Thick Coaxial 500 meters
10BaseF Fiber Optic 2000 meters
Wireless LANs
Not all networks are connected with cabling; some networks are wireless. Wireless LANs
use high frequency radio signals or infrared light beams to communicate between the
workstations and the file server. Each workstation and file server on a wireless network
has some sort of transceiver/antenna to send and receive the data. Information is relayed
between transceivers as if they were physically connected. For longer distance, wireless
communications can also take place through cellular telephone technology or by satellite.

Wireless networks are great for allowing laptop computers or remote computers to
connect to the LAN. Wireless networks are also beneficial in older buildings where it
may be difficult or impossible to install cables.

Wireless LANs also have some disadvantages. They are very expensive, provide poor
security, and are susceptible to electrical interference from lights and radios. They are
also slower than LANs using cabling.

Installing Cable - Some Guidelines

When running cable, it is best to follow a few simple rules:

• Always use more cable than you need. Leave plenty of slack.
• Test every part of a network as you install it. Even if it is brand new, it may have
problems that will be difficult to isolate later.
• Stay at least 3 feet away from fluorescent light boxes and other sources of
electrical interference.
• If it is necessary to run cable across the floor, cover the cable with cable
protectors.
• Label both ends of each cable.
• Use cable ties (not tape) to keep cables in the same location together.

Bridge - Devices that connect and pass packets between two network segments that use
the same communications protocol.

Multiplexer - A device that allows multiple logical signals to be transmitted

simultaneously across a single physical channel.

Network Modem - A modem connected to a Local Area Network (LAN) that is

accessible from any workstation on the network.

Network Interface Card (NIC) - A board that provides network communication

capabilities to and from a computer.
Ports - A connection point for a cable.

Repeater - A device used in a network to strengthen a signal as it is passed along the

network cable.

RJ-45 - Standard connectors used for unshielded twisted-pair cable.

Router -A device that routes information between interconnected networks. It can select
the best path to route a message, as well as translate information from one network to
another. It is similar to a superintelligent bridge.

Hub
An Ethernet hub or concentrator is a device for connecting multiple twisted pair or fiber
optic Ethernet devices together, making them act as a single segment. It works at the
physical layer of the OSI model, repeating the signal that comes into one port out each of
the other ports. If a signal comes into two ports at the same time a collision occurs, so
every attached device shares the same collision domain. Hubs support only half duplex
Ethernet, providing bandwidth which is shared among all the connected devices.

Most hubs detect typical problems such as excessive collisions on individual ports, and
partition the port, disconnecting it from the shared medium. Thus hub-based Ethernet is
generally more robust than coaxial cable based Ethernet where a misbehaving device can
disable the entire segment. Even if not partitioned automatically, a hub makes
troubleshooting easier because status lights can indicate the possible problem source or,
as a last resort, devices can be disconnected from a hub one at a time much more easily
than a coaxial cable.

Although switches are much more common, hubs are still useful in special
circumstances:-A protocol analyzer connected to a switch does not always receive all the
desired packets since the switch separates the ports into different segments. Connecting it
to a hub allows it to see all the traffic going through the hub.

Some computer clusters require each member computer to receive all of the traffic going
to the cluster. A hub will do this naturally; using a switch requires implementing special
tricks.

When a switch is accessible for end users to make connections (for example, in a
conference room), an inexperienced or careless user (or saboteur) can bring down the
network by connecting two ports together, causing a loop. This can be prevented by using
a hub, where a loop will break other users on the hub but not the rest of the network.
Switch
A switch is a device for making or breaking an electric circuit, or for selecting between
multiple circuits. In the simplest case, a switch has two pieces of metal called contacts
that touch to make a circuit, and separate to break the circuit. The contact material is
chosen for its resistance to corrosion, because most metals form insulating oxides that
would prevent the switch from working. Sometimes the contacts are plated with noble
metals. They may be designed to wipe against each other to clean off any contamination.
Nonmetallic conductors, such as conductive plastic, are sometimes used. The moving part
that applies the operating force to the contacts is called the actuator, and may be a toggle
or dolly, a rocker, a push-button or any type of mechanical linkage

A pair of contacts is said to be 'closed' when there is no space between them, allowing
electricity to flow from one to the other. When the contacts are separated by a space, they
are said to be 'open', and no electricity can flow. Switches can be classified according to
the arrangement of their contacts. Some contacts are normally open until closed by
operation of the switch, while others are normally closed and opened by the switch
action. A switch with both types of contact is called a changeover switch. The terms pole
and throw are used to describe switch contacts. A pole is a set of contacts that belong to a
single circuit. A throw is one of two or more positions that the switch can adopt. These
terms give rise to abbreviations for the types of switch which are used in the electronics
industry. In mains wiring names generally involving the word way are used; however,
these terms differ between British and American English and the terms two way and three
way are used in both with different meanings. Switches with larger numbers of poles or
throws can be described by replacing the "S" or "D" with a number or in some cases the
letter T (for triple). In the rest of this article the terms SPST SPDT and intermediate will
be used to avoid the ambiguity in the use of the word "way".

In a multi-throw switch, there are two possible transient behaviors as you move from one
postion to another. In some switch designs, the new contact is made before the old
contact is broken. This is known as make-before-break, and ensures that the moving
contact never sees an open circuit. The alternative is break-before-make, where the old
contact is broken before the new one is made. This ensures that the two contacts are
never shorted to each other. Both types of design are in common use, for different
applications

A biased switch is one containing a spring that returns the actuator to a certain position.
The "on-off" notation can be modified by placing parentheses around all positions other
than the resting position. For example, an (on)-off-(on) switch can be switched on by
moving the actuator in either direction away from the centre, but returns to the central off
position when the actuator is released.The momentary push-button switch is a type of
biased switch. The most common type is a push-to-make switch, which makes contact
when the button is pressed and breaks when the button is released. A push-to-break
switch, on the other hand, breaks contact when the button is pressed and makes contact
when it is released. An example of a push-to-break switch is a button used to release a
door held open by an electromagnet. Changeover push button switches do exist but are
even less common.

Net work Cables and Stuff:

In the network you will commonly find three types of cables used these are the, coaxial
cable, fiber optic and twisted pair.
Thick Coaxial Cable
This type cable is usually yellow in color and used in what is called thicknets, and has
two conductors. This coax can be used in 500-meter lengths. The cable itself is made up
of a solid center wire with a braided metal shield and plastic sheathing protecting the rest
of the wire.
Thin Coaxial Cable
As with the thick coaxial cable is used in thicknets the thin version is used in thinnets.
This type cable is also used called or referred to as RG-58. The cable is really just a
cheaper version of the thick cable.
Fiber Optic Cable
As we all know fiber optics are pretty darn cool and not cheap. This cable is smaller and
can carry a vast amount of information fast and over long distances.
Twisted Pair Cables
These come in two flavors of unshielded and shielded
Shielded Twisted Pair (STP)

Is more common in high-speed networks. The biggest difference you will see in the UTP
and STP is that the STP use's metallic shield wrapping to protect the wire from
interference.

-Something else to note about these cables is that they are defined in numbers also. The
bigger the number the better the protection from interference. Most networks should go
with no less than a CAT 3 and CAT 5 is most recommended.

-Now you know about cables we need to know about connectors. This is pretty important
and you will most likely need the RJ-45 connector. This is the cousin of the phone jack
connector and looks real similar with the exception that the RJ-45 is bigger. Most
commonly your connector are in two flavors and this is BNC (Bayonet Naur Connector)
used in thicknets and the RJ-45 used in smaller networks using UTP/STP.
Unshielded Twisted Pair (UTP)
This is the most popular form of cables in the network and the cheapest form that you can
go with. The UTP has four pairs of wires and all inside plastic sheathing. The biggest
reason that we call it Twisted Pair is to protect the wires from interference from
themselves. Each wire is only protected with a thin plastic sheath.
Ethernet Cabling
Now to familiarize you with more on the Ethernet and it's cabling we need to look at the
10's. 10Base2, is considered the thin Ethernet, thinnet, and thinwire which uses light
coaxial cable to create a 10 Mbps network. The cable segments in this network can't be
over 185 meters in length. These cables connect with the BNC connector. Also as a note
these unused connection must have a terminator, which will be a 50-ohm terminator.

10Base5, this is considered a thicknet and is used with coaxial cable arrangement such as
the BNC connector. The good side to the coaxial cable is the high-speed transfer and
cable segments can be up to 500 meters between nodes/workstations. You will typically
see the same speed as the 10Base2 but larger cable lengths for more versatility.

10BaseT, the “T” stands for twisted as in UTP (Unshielded Twisted Pair) and uses this
for 10Mbps of transfer. The down side to this is you can only have cable lengths of 100
meters between nodes/workstations. The good side to this network is they are easy to set
up and cheap! This is why they are so common an ideal for small offices or homes.

100BaseT, is considered Fast Ethernet uses STP (Shielded Twisted Pair) reaching data
transfer of 100Mbps. This system is a little more expensive but still remains popular as
the 10BaseT and cheaper than most other type networks. This on of course would be the
cheap fast version.

10BaseF, this little guy has the advantage of fiber optics and the F stands for just that.
This arrangement is a little more complicated and uses special connectors and NIC's
along with hubs to create its network. Pretty darn neat and not to cheap on the wallet.

An important part of designing and installing an Ethernet is selecting the appropriate

Ethernet medium. There are four major types of media in use today: Thickwire for
10BASE5 networks, thin coax for 10BASE2 networks, unshielded twisted pair (UTP) for
10BASE-T networks and fiber optic for 10BASE-FL or Fiber-Optic Inter-Repeater Link
(FOIRL) networks. This wide variety of media reflects the evolution of Ethernet and also
points to the technology's flexibility. Thickwire was one of the first cabling systems used
in Ethernet but was expensive and difficult to use. This evolved to thin coax, which is
easier to work with and less expensive.

Network Interface Cards:

Network interface cards, commonly referred to as NICs, and are used to connect a PC to
a network. The NIC provides a physical connection between the networking cable and the
computer's internal bus. Different computers have different bus architectures; PCI bus
master slots are most commonly found on 486/Pentium PCs and ISA expansion slots are
commonly found on 386 and older PCs. NICs come in three basic varieties: 8-bit, 16-bit,
and 32-bit. The larger the number of bits that can be transferred to the NIC, the faster the
NIC can transfer data to the network cable.

Many NIC adapters comply with Plug-n-Play specifications. On these systems, NICs are
automatically configured without user intervention, while on non-Plug-n-Play systems,
configuration is done manually through a setup program and/or DIP switches.

Cards are available to support almost all networking standards, including the latest Fast
Ethernet environment. Fast Ethernet NICs are often 10/100 capable, and will
automatically set to the appropriate speed. Full duplex networking is another option,
where a dedicated connection to a switch allows a NIC to operate at twice the speed.
Hubs/Repeaters:
Hubs/repeaters are used to connect together two or more Ethernet segments of any media
type. In larger designs, signal quality begins to deteriorate as segments exceed their
maximum length. Hubs provide the signal amplification required to allow a segment to be
extended a greater distance. A hub takes any incoming signal and repeats it out all ports.

Ethernet hubs are necessary in star topologies such as 10BASE-T. A multi-port twisted
pair hub allows several point-to-point segments to be joined into one network. One end of
the point-to-point link is attached to the hub and the other is attached to the computer. If
the hub is attached to a backbone, then all computers at the end of the twisted pair
segments can communicate with all the hosts on the backbone. The number and type of
hubs in any one-collision domain is limited by the Ethernet rules. These repeater rules are
discussed in more detail later.
 Max Nodes Max Distance
Network Type
Per Segment Per Segment
10BASE-T 2 100m
10BASE2 30 185m
10BASE5 100 500m
10BASE-FL 2 2000m

Adding Speed:
While repeaters allow LANs to extend beyond normal distance limitations, they still limit
the number of nodes that can be supported. Bridges and switches, however, allow LANs
to grow significantly larger by virtue of their ability to support full Ethernet segments on
each port. Additionally, bridges and switches selectively filter network traffic to only
those packets needed on each segment - this significantly increases throughput on each
segment and on the overall network. By providing better performance and more
flexibility for network topologies, bridges and switches will continue to gain popularity
among network managers.
Bridges:
The function of a bridge is to connect separate networks together. Bridges connect
different networks types (such as Ethernet and Fast Ethernet) or networks of the same
type. Bridges map the Ethernet addresses of the nodes residing on each network segment
and allow only necessary traffic to pass through the bridge. When a packet is received by
the bridge, the bridge determines the destination and source segments. If the segments are
the same, the packet is dropped ("filtered"); if the segments are different, then the packet
is "forwarded" to the correct segment. Additionally, bridges do not forward bad or
misaligned packets.

Bridges are also called "store-and-forward" devices because they look at the whole
Ethernet packet before making filtering or forwarding decisions. Filtering packets, and
regenerating forwarded packets enable bridging technology to split a network into
separate collision domains. This allows for greater distances and more repeaters to be
used in the total network design.
Ethernet Switches:
Ethernet switches are an expansion of the concept in Ethernet bridging. LAN switches
can link four, six, ten or more networks together, and have two basic architectures: cut-
through and store-and-forward. In the past, cut-through switches were faster because they
examined the packet destination address only before forwarding it on to its destination
segment. A store-and-forward switch, on the other hand, accepts and analyzes the entire
packet before forwarding it to its destination.

It takes more time to examine the entire packet, but it allows the switch to catch certain
packet errors and keep them from propagating through the network. Both cut-through and
store-and-forward switches separate a network into collision domains, allowing network
design rules to be extended. Each of the segments attached to an Ethernet switch has a
full 10 Mbps of bandwidth shared by fewer users, which results in better performance (as
opposed to hubs that only allow bandwidth sharing from a single Ethernet). Newer
switches today offer high-speed links, FDDI, Fast Ethernet or ATM. These are used to
link switches together or give added bandwidth to high-traffic servers. A network
composed of a number of switches linked together via uplinks is termed a "collapsed
backbone" network.
Routers:
Routers filter out network traffic by specific protocol rather than by packet address.
Routers also divide networks logically instead of physically. An IP router can divide a
network into various subnets so that only traffic destined for particular IP addresses can
pass between segments. Network speed often decreases due to this type of intelligent
forwarding. Such filtering takes more time than that exercised in a switch or bridge,
which only looks at the Ethernet address. However, in more complex networks, overall
efficiency is improved by using routers.

Types of Servers:
Device Servers
A device server is defined as a specialized, network-based hardware device designed to
perform a single or specialized set of server functions. It is characterized by a minimal
operating architecture that requires no per seat network operating system license, and
client access that is independent of any operating system or proprietary protocol. In
addition the device server is a "closed box," delivering extreme ease of installation,
minimal maintenance, and can be managed by the client remotely via a Web browser.

Print servers, terminal servers, remote access servers and network time servers are
examples of device servers which are specialized for particular functions. Each of these
types of servers has unique configuration attributes in hardware or software that help
them to perform best in their particular arena.
Print Servers
Print servers allow printers to be shared by other users on the network. Supporting either
parallel and/or serial interfaces, a print server accepts print jobs from any person on the
network using supported protocols and manages those jobs on each appropriate printer.

Print servers generally do not contain a large amount of memory; printers simply store
information in a queue. When the desired printer becomes available, they allow the host
to transmit the data to the appropriate printer port on the server. The print server can then
simply queue and print each job in the order in which print requests are received,
regardless of protocol used or the size of the job.
Multiport Device Servers
Devices that are attached to a network through a multiport device server can be shared
between terminals and hosts at both the local site and throughout the network. A single
terminal may be connected to several hosts at the same time (in multiple concurrent
sessions), and can switch between them. Multiport device servers are also used to
network devices that have only serial outputs. A connection between serial ports on
different servers is opened, allowing data to move between the two devices.

Given its natural translation ability, a multi-protocol multiport device server can perform
conversions between the protocols it knows, like LAT and TCP/IP. While server
bandwidth is not adequate for large file transfers, it can easily handle host-to-host
inquiry/response applications, electronic mailbox checking, etc. And it is far more
economical than the alternatives of acquiring expensive host software and special-
purpose converters. Multiport device and print servers give their users greater flexibility
in configuring and managing their networks.

Whether it is moving printers and other peripherals from one network to another,
expanding the dimensions of interoperability or preparing for growth, multiport device
servers can fulfill your needs, all without major rewiring.
Access Servers
While Ethernet is limited to a geographic area, remote users such as traveling sales
people need access to network-based resources. Remote LAN access, or remote access, is
a popular way to provide this connectivity. Access servers use telephone services to link
a user or office with an office network. Dial-up remote access solutions such as ISDN or
asynchronous dial introduce more flexibility. Dial-up remote access offers both the
remote office and the remote user the economy and flexibility of "pay as you go"
telephone services. ISDN is a special telephone service that offers three channels, two 64
Kbps "B" channels for user data and a "D" channel for setting up the connection. With
ISDN, the B channels can be combined for double bandwidth or separated for different
applications or users. With asynchronous remote access, regular telephone lines are
combined with modems and remote access servers to allow users and networks to dial
anywhere in the world and have data access. Remote access servers provide connection
points for both dial-in and dial-out applications on the network to which they are
attached. These hybrid devices route and filter protocols and offer other services such as
modem pooling and terminal/printer services. For the remote PC user, one can connect
from any available telephone jack (RJ45), including those in a hotel rooms or on most
airplanes.
Network Time Servers
A network time server is a server specialized in the handling of timing information from
sources such as satellites or radio broadcasts and is capable of providing this timing data
to its attached network. Specialized protocols such as NTP or udp/time allow a time
server to communicate to other network nodes ensuring that activities that must be
coordinated according to their time of execution are synchronized correctly. GPS
satellites are one source of information that can allow global installations to achieve
constant timing.