-
What
is a Computer Network?
-
What
are the Benefits of Computer Networking?
-
Hardware
Technology
-
Connecting
Network Devices
-
Controlling
Data Transmission
-
The
Network Operating System
-
Desktop
Operating Systems
-
Application
Software
-
Internetworking
-
Real
World Networking
-
Host
Access
-
Wide
Area Networking
-
Important
High-Speed and WAN Technologies
-
Global
Networking
-
Primer
Appendix
What Is a Computer
Network?
Commonly connected
devices include microcomputers, minicomputers, mainframe computers, terminals,
printers, fax machines, pagers, and various data storage devices. In the
near future, numerous other types of devices will be network connectable,
including interactive TVs, videophones, and navigational and environmental
control systems. Eventually, devices everywhere will give you two-way access
to a vast array of resources on a global computer network.
In today's business
world, a computer network is much more than a collection of interconnected
devices. For many businesses, the computer network is the resource that
enables them to gather, analyze, organize, and disseminate the information
that is essential to their profitability. The rise of intranets and extranets—a
recent development in computer networking—is the latest indication of the
crucial importance of computer networking to businesses. Intranets and
extranets are private business networks that are based on Internet technology.
Intranets, extranets, and the Internet will be treated in more detail in
a later section. For now, it is enough to understand that businesses are
currently implementing intranets at a breakneck pace and for one reason
only—an intranet enables a business to collect, manage, and disseminate
information more quickly and easily than ever before. Many businesses are
implementing intranets simply to remain competitive; businesses that delay
are likely to see their competition outdistance them.
What Are the Benefits
of Computer Networking?
In addition to information
storage and retrieval, there is a host of other important benefits of networking
computers. Having a computer network enables you to combine the skills
of different people and the power of different equipment, regardless of
the physical locations of the people or the equipment. And computer networking
enables people to easily share information, allowing them to work more
securely, efficiently, and productively.
Powerful, Flexible Collaboration
For example, a managing
editor, associate editors, writers, and artists may need to work together
on a publication. With a computer network, they can share the same electronic
files, each from his or her own computer, without copying or transferring
files. If the applications they are using feature even basic integration
with the network operating system, they can perform such tasks as opening,
viewing, and printing the same file simultaneously.
Using applications
that are designed to take full advantage of network capabilities and services,
network users can collaborate with ease and speed. For example, users can
engage in real-time teleconferencing, talking face-to-face while simultaneously
viewing and editing the same document, adding and deleting notes and comments,
and instantaneously viewing each other's changes as they are made. And,
they can do this without having to worry about accidentally changing or
deleting the work of others.
To be able to collaborate
electronically from widely separate physical locations has significant
advantages. It enables people to avoid the considerable time investments
and costs connected with traveling. It enables people to communicate instantaneously,
regardless of the distance, and to act before their competitors do. It
frees people from having to reconcile the differences in multiple information
files. Electronic collaboration enables people to minimize the amount of
work required to complete projects—it frees them from redoing work they
would do correctly in the first place if they had instantaneous access
to up-to-date information and instructions.
Freedom to Choose the
Right Tool
The design of any particular
computer can make it well suited for some tasks and not as well suited
for others. In an open environment, you can combine many kinds of computers
to take advantage of the special strengths of each type of machine. For
example, Novell network users can use IBM PCs running any version of Windows
or DOS, Macintosh computers running a version of the Macintosh operating
system, Sun workstations running the UNIX operating system, and many other
types of computers, all on the same network. Scientists, secretaries, doctors,
lawyers, writers, editors, artists, engineers—everyone can use the type
of computer equipment best suited to the type of work he or she does, yet
each can still easily share information with everyone else.
Cost-Effective Resource
Sharing
Equipment sharing has
significant benefits. It enables you to buy equipment with features that
you wouldn't otherwise be able to afford and to ensure that the equipment
is used to its full potential. A correctly implemented network can result
in both increased productivity and lower equipment costs.
For example, suppose
you had a number of unconnected computers. People using these computers
would not be able to print unless you purchased a printer for each computer
or unless users manually transferred files from computers without printers
to those with printers. In choosing between these alternatives, you would
be choosing between significant expenses for hardware or significant expenses
for labor.
But networking the
computers would give you other alternatives. Because all users could share
any networked printer, you would not need to buy a printer for every computer.
Therefore, rather than buying numerous inexpensive printers, none of which
had top-end productivity features and all of which would sit idle most
of the time, you could buy a few inexpensive printers and a few printers
with top-end productivity features. The more powerful printers might be
able to print 20 times more pages per minute than the inexpensive printers.
And, the more powerful printers might also be able to print in color and
to sort, staple, or bind any number of pages, and to produce large numbers
of completed documents.
On a Novell network,
all users could share the various printers, accessing whichever printer
was most appropriate for the job they were doing. The network software
would enable users to print whenever they wanted. The network would print
documents in the order they were received, on the printer the user selected.
Whenever necessary, users would be able to change the order in which documents
were to be printed and where they were to be printed.
By selecting the right
mix of printers and allowing each network user appropriate access to them,
you could have enough printing power to take care of the needs of all users;
you could ensure that expensive equipment was not standing idle; and you
could provide users with the latest, most powerful productivity features,
freeing them from many tasks they would otherwise have to do manually—all
for a significantly lower cost than if you were to buy an inexpensive printer
for each of the computers connected to your network.
A network enables you
to share any networkable equipment or software and realize the same benefits
that you would enjoy from sharing printers. On a network, users can share
modems; data storage devices, such as hard disks and CD-ROM drives; data
backup devices, such as tape drives; E-mail systems; facsimile machines;
and all networkable software. When you compare sharing these resources
to purchasing them for each computer, the cost savings can be enormous.
When you implement
an intranet, you can share network resources with suppliers, consultants,
and other outside partners. Soon, you will be able to allow your employees
to rent applications over the Internet. Businesses have just begun to explore
the possibilities of intranet resource sharing.
Secure Management of Sensitive
Information
Effective Worldwide Communications
Easy, Immediate Information
Dissemination
Worldwide, Instantaneous
Access to Information
Integrated, flexible
information sharing; instantaneous information updating and access; lower
equipment costs; flexible use of computing power; secure management of
sensitive information—these are the benefits of computer networking. And
these benefits help us produce the results we are all looking for: increased
efficiency, productivity, and profitability.
Hardware
Technology
Definition: Data Versus
Information
Data are entities that
convey meaning. Computer data is stored as a series of electrical charges
arranged in patterns to represent information. In other words, data refers
to the form of the information (the electrical patterns). It is not the
information itself.
Information means decoded
data, in human-readable form. In other words, information is the real-world,
useful form of data. For example, the data in an electronic file can be
decoded and displayed on a computer screen or printed onto paper as a business
letter.
Encoding and Decoding
Data
To encode information
into data and later decode that data back into information, we use electronic
devices (one of which is the computer) that generate electronic signals.
Signals are simply the electric or electromagnetic encoding of data. Various
components in a computer enable it to generate signals to perform the encoding
and decoding task.
How Network Data Is Transferred
Let's suppose we've
agreed on a coding scheme and we have several computers that are all capable
of encoding and decoding the information we want to save.
Now, to have a computer
network, we need only the means of transferring the generated signals between
the computers. To transfer signals between the computers, we need two things:
(1) a transmission medium to carry the signals and (2) devices to propagate
(send) and receive the signals.
Network Transmission Media
There are two types
of transmission media: guided media and unguided media.
Guided media are manufactured
so that signals will be confined to a narrow path and will behave predictably.
Commonly used guided media include twisted-pair wiring, similar to common
telephone wiring; coaxial cable, similar to that used for cable TV; and
optical fiber cable.
Figure
1: Common guided transmission media
Unguided media are
natural parts of the existing environment that can be used as physical
paths to carry electrical signals. Earth's atmosphere and outer space are
examples of unguided media that are commonly used to carry signals. These
media can carry such electromagnetic signals as microwave and infrared
light waves.
Regardless of the type
of medium, the network signal is transmitted through it as some kind of
waveform. When transmitted through wire and cable, the signal is an electrical
waveform. When transmitted through fiber-optic cable, the signal is a light
wave, somewhere in the spectrum of visible or infrared light. When transmitted
through Earth's atmosphere or outer space, the signal can take the form
of waves in the radio spectrum, including VHF and microwaves, or it can
be light waves, including infrared or visible light (for example, lasers).
When planning a computer
network, designers choose a transmission medium, or a combination of media,
based on the physical circumstances involved in building the network and
the reliability and data-handling performance required of the network.
The objective is to keep costs to a minimum yet provide all parts of the
network with the required reliability and performance.
For example, if you
needed to build a network consisting of two subnetworks located in separate
buildings several miles apart, you might use two or more transmission media.
If you did not require the same level of performance on both subnetworks,
you might use a different type of wire or cable as the transmission medium
on each.
To connect the two
subnetworks across town and ensure a reliable connection even in rain and
fog, you might use a third medium, Earth's atmosphere, and connect the
subnetworks through a microwave link. Or, you might use a T1 or T3 connection.
T1 and T3 are dedicated lines (basically special telephone lines) that
support high-speed communications. They can be leased from private companies
that specialize in providing communication services.
Transmitting and Receiving
Devices
Network Adapters
Network adapters are
manufactured for connection to virtually any type of guided medium, including
twisted-pair wire, coaxial cable, and fiber-optic cable. They are also
manufactured for connection to devices that transmit and receive visible
light, infrared light, and radio microwaves, to enable wireless networking
across the unguided media of Earth's atmosphere and outer space.
The connection hardware
used to make connections between network adapters and different transmission
media depends on the type of medium used. For example, twist-on BNC connectors
are commonly used for connection to coaxial cable, while snap-in telephone-type
jacks are ordinarily used for connection to twisted-pair wiring. Figure
2 shows two different types of network adapters connected to different
computers and media, using different types of connectors.
Figure
2: Network adapters are manufactured in a variety of forms, for virtually
every kind of communication medium.
Repeaters
As a signal travels
through a transmission medium, it encounters resistance and gradually becomes
weak and distorted. The technical term for this signal weakening is "attenuation."
All signals attenuate, and at some point they become too weak and distorted
to be reliably received. Repeaters are used to overcome this problem.
A simple, dedicated
repeater is a device that receives the network signal and retransmits it
at the original transmission strength. Repeaters are placed between other
transmitting and receiving devices on the transmission medium, at a point
where the signal will not have attenuated too much to be reliably received.
In today's networks,
dedicated repeaters are seldom used. Repeating capabilities are built into
other, more complex networking devices. For example, virtually all modern
network adapters, hubs, and switches incorporate repeating capabilities.
Wiring Concentrators,
Hubs, and Switches
In most cases, hubs,
wiring concentrators, and switches are proprietary, standalone hardware.
There are a number of companies that manufacture such equipment. Occasionally,
hub technology consists of hub cards and software that work together in
a standard computer.
Figure 3 shows two
common hardware-based connection devices: a token-ring switch and an Ethernet
10Base-T concentrator.
Figure
3: Token-ring switch and Ethernet 10Base-T concentrator
Microwave Transmitters
Figure 4: Satellite microwave link
Infrared and Laser Transmitters
Modems
A transmitting modem
converts (modulates) the encoded data signal to an audible signal and transmits
it. A modem connected at the other end of the line listens to the audible
signal and converts it back into a digital signal (demodulates it) for
the computer on the receiving end of the communication link. Modems are
commonly used for inexpensive, intermittent communications between geographically
isolated computers and a main network.
Connecting
Network Devices
First, we will look
at how guided transmission media are commonly connected physically to form
the physical topology of a local area network. Then, we will examine three
logical topologies, the electronic schemes used to connect network devices.
Common Physical Topologies:
Bus, Star, and Star-Wired Ring
Physical Bus
Figure
5: Physical bus topology
A more complex form
of the physical bus topology is the distributed bus (also called the tree
topology). In the distributed bus, the trunk cable starts at what is called
a "root," or "head end," and branches at various points along the way.
(Thus, unlike the simple bus topology described above, this variation uses
a trunk cable with more than two end points.) Where the trunk cable branches,
the division is made by means of a simple connector (as opposed to the
star physical topology discussed below, where connections are made to a
central, somewhat sophisticated connection device). The distributed bus
topology is illustrated in Figure 6.
Figure
6: Distributed bus topology
Physical Star
In a real-life implementation
of even a simple physical star topology, the actual layout of the transmission
media need not form a recognizable star pattern; the only required physical
characteristic is that each network device be connected by its own cable
to the central connection point.
The simplest form of
the physical star topology is illustrated in Figure 7.
Figure
7: Physical star topology
A more complex form
of the physical star topology is the distributed star. In this topology,
there are multiple central connection points, which are all connected to
form a string of stars. This topology is illustrated in Figure 8.
Figure
8: Distributed star topology
Physical Star-Wired Ring
Figure
9: Physical star-wired ring topology
In the star-wired ring
physical topology, the hubs are "intelligent." If the physical ring is
somehow broken, each hub is able to close the physical circuit at any point
in its internal ring so that the ring is restored. Refer to details shown
in Figure 9, hub A, to see how this works.
Currently, the star
topology and its derivatives are most preferred by network designers and
installers because using these topologies makes it simple to add network
devices anywhere. In most cases, you can simply install one new cable between
the central connection point and the desired location of the new network
device, without moving or adding to a trunk cable or making the network
unavailable for use by other stations.
Common Logical Topologies:
Bus, Ring, and Star (Switching)
There are three basic
logical topologies, each of which has distinct advantages in specific situations.
As you study the figures representing these topologies, remember that the
figures represent a logical (electronic), not a physical, connection scheme.
Logical Bus
On a logical bus network,
the transmission media is shared. To prevent transmission interference,
only one station may transmit at a time. Thus, there must be a method for
determining when each station is allowed to use the media. This method
is called the media access control (MAC).
The media access control
method most commonly used for a logical bus network is a contention method
called "carrier sense multiple access with collision detection (CSMA/CD)."
This media access control method is similar to the access scheme used on
a telephone party line. When any station wants to send a transmission,
it "listens" (carrier sense) to determine if another station is currently
transmitting on the media. If another station is transmitting, the station
that wants to transmit waits. When the media become free, the waiting station
transmits. If two or more stations determine that the media are free and
transmit simultaneously, there is a "collision." All transmitting stations
detect the collision, transmit a brief signal to inform all other stations
there has been a collision, and all stations then wait a random amount
of time before attempting to transmit.
A logical bus network
may also use token passing for media access control. In this MAC method,
each network station is assigned a logical position in an ordered sequence,
with the last number of the sequence pointing back to the first (the logical
order that the stations are assigned need not correspond with any physical
order). A control frame, called a "token," is used to control which station
can use the media. A station can transmit only when in possession of the
token. Furthermore, a station can have the token only a limited time before
it must pass the token to the next station. The token starts at the first
station in the predefined logical order. While the first station has the
token, it transmits, polls stations, and receives responses (gives other
stations permission to use the media) until the allotted time expires;
or, it passes the token when it no longer needs control of the media, whichever
happens first. The first station passes the token to the second station
in the logical sequence. This token passing (in sequence) continues nonstop
while the network is running—thus, every station gets equitable access
to the transmission media.
The logical bus transmission
scheme is used in combination with both the physical bus and physical star
topology, and the MAC method can vary in different cases. For example,
the cable on thin Ethernet networks is laid out as a physical bus and the
transmission scheme is a logical bus, but the cable on 10Base-T Ethernet
networks and on ARCnet networks is laid out as a physical star, although
both use the logical bus transmission scheme. And thin Ethernet (physical
bus) and 10Base-T Ethernet (physical star) both use the CSMA/CD MAC method,
but ARCnet (physical star) uses token passing as its MAC method.
Figure 10 shows a thin
Ethernet network (physical bus, logical bus), and Figure 11 shows a 10Base-T
Ethernet network (physical star, logical bus). In both figures, notice
that the network signal (shown by the arrows) emanates from the sending
station and travels in all directions, to all parts of the transmission
media (the determining criterion for a logical bus topology).
Figure
10: Thin Ethernet network (physical bus, logical bus)
Figure
11: 10Base-T Ethernet network (physical star, logical bus)
Logical Ring
Figure
12: Logical ring topology
Media access control
for the logical ring topology is almost always based on a form of token
passing, the basics of which are described in the logical bus topology
section. (Stations are not necessarily granted media access in the same
order in which they receive frames on the physical ring.) IBM's Token-Ring
network is a logical ring network based on the star-wired ring physical
topology.
Logical Star (Switching)
In its pure form, switching
provides a dedicated line for each end station. This means that when one
station transmits a signal to another station on the same switch, the switch
transmits the signal only on the two paths connecting the sending and receiving
station. Figure 13 shows how data would be transmitted from one station
to another if two stations were directly connected to the same switch.
Figure 13: Switching
Most switching technology
adds switching capability to existing connection standards, incorporating
the logical connection schemes (including the media access control methods)
of the existing standards.
For example, a 10Base-T
Ethernet switch supports the Ethernet CSMA/CD media access control method.
Some switches are designed to support and combine multiple network standards.
For example, a switch might contain both 10Base-T Ethernet ports and Fiber
Distributed Data Interface (FDDI) ports. In this case, the switch would
support the logical connection scheme for both standards, including the
Ethernet CSMA/CD and the FDDI token-ring MAC methods.
Switches have built-in
connection logic and significant amounts of fast memory. This enables them
to simultaneously service all connected stations at full access speed.
Thus, when you connect a station directly to a switch, you can increase
the total throughput of your network—a significant performance advantage.
Switching illustrates
well that a logical topology consists of the total of the various aspects
of the electronic connection scheme, not just the MAC method. By combining
new (switching) capabilities with existing logical connection schemes,
engineers create a new logical topology.
Switching can be distributed
(multiple switches can be connected using one or more physical topologies).
Switches can be used not only to connect individual stations, but also
to connect network segments (groups of stations). Thus, in many circumstances,
switching can be used to improve the performance of your network.
Connecting a Simple Network
Figure
14: Various networking hardware connected to form a simple network
The network in this
illustration includes the following components: three computers connected
through a 10Base-T concentrator by means of unshielded twisted-pair wiring;
three Ethernet 10Base-T network adapters, one installed inside each of
the computers; and a laser printer that is connected to one of the computers.
The computer at the
bottom center of the illustration is a network server; it controls the
network (details will be covered in a following section). The other two
computers are workstations. The workstations use the network under the
control of the network server. One workstation is an IBM PC and the other
is an Apple Macintosh computer.
The 10Base-T concentrator
serves as a common connection point for the three computers; it repeats
network signals.
The lines between the
different components of the network represent the transmission medium,
which is twisted-pair wiring. As you may remember from our recent discussion
of topologies, this 10Base-T network is connected in a physical star, but
it is based on a logical bus that uses a contention scheme as the means
for workstations to get access to the transmission medium.
The printer in this
network is connected directly to the server by means of a parallel interface
cable, which is a standard connection method. The server accepts print
jobs from either workstation and sends the jobs through the parallel interface
cable to the printer. This is the simplest way to enable both workstations
to use the printer. There are other ways to connect printers to a network,
including attaching them to a computer set up as a dedicated print server
or connecting them to a computer that runs special software enabling it
to function as both a workstation and a print server. Many printers are
now manufactured with an internal network adapter so that they can be attached
directly to the transmission medium at any physical point in the network.
Controlling Data
Transmission
But simply connecting
hardware doesn't make a computer network. Even though the hardware is capable
of generating signals and transmitting them across a medium, it must be
told when and how to do this. There must be network communication software
to tell the hardware when and how to transmit. The software and hardware
on all parts of the network must work together to enable the transmission
of data from one networked computer to another. We'll explore various networking
software a little later. First, let's look at the communication model that
is the basis for controlling data transmission on computer networks.
To guarantee reliable
transmission of data, there must be an agreed method that governs how data
is sent and received. For example, how does a sending computer indicate
which computer it is sending data to? And, if the data will be passed through
intervening devices, how are these devices to understand how to handle
the data so that it will get to the intended destination? And, what if
the sending and receiving computers use different data formats and data
exchange conventions—how will data be translated to allow its exchange?
These are only a few of the questions that must be answered before data
can be reliably transmitted and received across a computer network.
Understanding the Open
Systems Interconnection (OSI) model will allow you to understand how data
can be transferred between two networked computers, regardless of whether
they are on the same network, or are the same type of computer, or use
the same data formats and exchange conventions.
ISO and the OSI Model
The OSI Model: What It
Is and Why It's Important
First, when a vendor's
products adhere to the standards the OSI model has spawned, connecting
those products to other vendors' products is relatively simple. Conversely,
the further a vendor departs from those standards, the more difficult it
becomes to connect that vendor's products to those of other vendors. Second,
if a vendor were to depart from the communication standards the model has
spawned, software development efforts would be very difficult because the
vendor would have to build every part of all necessary software, rather
than often being able to build on the existing work of other vendors.
The first two problems
give rise to a third significant problem for vendors: A vendor's products
become less marketable as they become more difficult to connect with other
vendors' products unless the introduction of the vendor's products is well
ahead of the introduction of other such products into the general marketplace.
Now, keeping in mind
the purpose of the OSI model, let's take a look at its structure.
The Seven Layers of the
OSI Model
Each layer of the OSI
model contains a logically grouped subset of the functions required for
controlling network communications. The seven layers of the OSI model and
the general purpose of each are shown in Figure 15.
Figure 15: The OSI model
Standards and Protocols
A standard for any
layer of the OSI model specifies the communication services to be provided
and a protocol that will be used as a means to provide those services.
A protocol is a set of rules network devices must follow (at any OSI layer)
to communicate. A protocol consists of the control functions, the control
codes, and the procedures necessary for successfully transferring data.
For every layer of
the OSI model, there is more than one protocol standard. This is because
a number of standards were proposed for each layer and because the various
organizations that defined those standards—specifically, the standards
committees inside these organizations—decided that more than one of the
proposed standards had real merits. Thus, they allowed for the use of different
standards to satisfy different networking needs.
Network Communications
Through the OSI Model
Figure
16: Networked computers communicating through the OSI model
Our figure represents
two networked computers, each of which is running various pieces of software
(most not shown). Running together, the various pieces of software implement
the seven OSI layers. These computers are identical: They are running identical
software, and they are using identical protocols at all OSI layers. Above
the OSI application layer, each computer is running an E-mail program.
The E-mail program enables the users of the two computers to exchange messages.
Our figure represents the transmission of one brief message from computer
A to computer B.
The transmission starts
with the user of computer A pressing a key to send a mail message to the
user of computer B. The E-mail application is designed to talk to the OSI
application layer—it knows the proper protocol for doing so. The E-mail
application transfers the message to the OSI application layer. Using the
functions built into its protocol, the application layer accepts the message
data and adds an application-layer header to it. The application-layer
header contains the information necessary for the application layer in
computer B to correctly handle the data when computer B receives it.
After adding its header,
the application layer in computer A passes the data to the presentation
layer below. The presentation layer treats everything received as data,
including the application-layer header, and appends its own header (the
technical term for this is "encapsulation"). The presentation-layer header
contains the information necessary for the presentation layer in computer
B to correctly handle the data. After adding its header, the presentation
layer transfers the new data unit to the session layer.
This process is repeated
through all layers in computer A until a final header is added at the data-link
layer. After the data-link-layer header is added, the data unit is known
as a "frame." The data, or frame, is passed from the data-link layer to
the physical layer and is transmitted across the transmission medium connecting
the two computers.
When the signal reaches
computer B, layer one in computer B (the physical layer) copies the data.
Now the process is reversed. The physical layer in computer B transfers
the data to the data-link layer. The data-link layer removes the header
information that was attached by the corresponding layer in computer A,
acts upon the information the header contains, and transfers the data unit
up to the network layer. This process continues, with the headers being
stripped off at each layer and the instructions contained therein carried
out, until the original data from computer A (the message) is finally passed
from the application layer to the E-mail application in computer B. When
the E-mail application receives the message, it displays the message on
the screen for the user of computer B to read.
Now look at Figure
16 and imagine what would be possible if the software implementing different
layers of the OSI model were able to handle not just one communication
protocol at any one layer, but almost any communication protocol used at
any layer, by any computer—there would be no limits to the interconnection
of dissimilar computing devices. This is the kind of power that will be
the basis for a global network—the networking of all kinds of business
and personal devices into the Information Superhighway. And this is the
kind of power built into NetWare® products.
Commonly Used Standards
To understand the capabilities
of NetWare products, it will help to know the OSI layer at which a particular
protocol operates and why the standard is important. As you shall see later,
by converting protocols or using multiple protocols at different layers
of the OSI model, it is possible to enable different computer systems to
share data, even if they use different software applications, operating
systems, and data-encoding techniques.
Figure 17 shows some
commonly used standards and the OSI layer at which they operate.
Figure
17: Important standards at various OSI layers
Layer-One Standards: Physical
Layer-Two Standards: Data
Link (Media Access Control and Logical Link Control)
Important recent technologies
at layer two include 100Base-T (IEEE 802.2u), 100VG-AnyLAN (802.12), and
Asynchronous Transfer Mode (ATM). The ATM standard is not yet fully defined.
Also, frame relay is an important layer-two WAN technology. These technologies
are treated in greater detail in a later section.
Layer-two standards
encompass two sublayers: media access control and logical link control.
Media Access Control
The IEEE 802.3 standard
specifies a medium-access method known as "carrier sense multiple access
with collision detection (CSMA/CD)." This medium-access method is the same
as the contention method described in the earlier discussion of topologies,
under the heading "Logical Bus."
The IEEE 802.4, 802.5,
and FDDI standards all specify some form of token passing as the media
access control method. The basics of the token-passing method were also
described earlier, also under the heading "Logical Bus."
In general, using a
form of token passing for the media access control works best when large
numbers of computers frequently send small amounts of data—for example,
when a number of workstations continually read and write small records
to and from a database. Contention schemes work well when computers send
large amounts of data intermittently—for example, during desktop publishing
or document imaging.
Logical Link Control
The IEEE 802.2 standard
is the most commonly used logical link control standard.
The Point-to-Point
Protocol (PPP) is an important standard at this OSI level. PPP is used
for communications across point-to-point links such as T1 and T3 lines.
It is an important protocol for wide area networking, which will be covered
later.
Layer-Three Standards:
Network
One important network-layer
standard is the Department of Defense (DOD) Internet Protocol (IP) specification,
which is part of the Transmission Control Protocol/Internet Protocol (TCP/IP)
standard developed by the DOD. This protocol has become extremely important
recently because it is the basis for the Internet and for all intranet
technology. Also, the Department of Defense will often not purchase networking
products that cannot communicate using this protocol.
Because Novell commands
a large share of the networking market, its native Internetwork Packet
Exchange™ (IPX) protocol, is also an important network-layer standard.
IPX is a connectionless datagram protocol. A connectionless protocol does
not need to establish a connection between two networked computers to transfer
information between them. Packet acknowledgment, or connection control,
is provided by protocols above IPX, such as Novell's Sequenced Packet Exchange™
(SPX). SPX will be explained in more detail in a later section. Because
IPX is a datagram protocol, each communication packet is treated as an
individual entity. IPX does not have to establish a logical or sequential
relation between packets. Thus, IPX is very efficient—it addresses and
transfers data with minimum control overhead.
IPX uses other NetWare
protocols that work at the network layer to accomplish internetwork routing.
These protocols, the Routing Information Protocol (RIP), the Service Advertising
Protocol (SAP), and the NetWare Link Services Protocol™ (NLSP), will be
explained in more detail in a later section.
The Consultative Committee
for International Telegraph and Telephone (CCITT) X.25 standard is another
commonly used network-layer standard. It specifies the interface for connecting
computers on different networks by means of an intermediate connection
made through a packet-switched network (for example, a common carrier network
such as CompuServe, Tymnet, or Telnet). The X.25 standard includes the
data-link and physical-layer protocols shown below it in Figure 17.
Apple Computer, Inc.
has established a set of protocols for its products, referred to collectively
as AppleTalk. At the network layer of the OSI model, the Apple protocol
is called Datagram Delivery Protocol. Figure 18 shows how the set of AppleTalk
protocols fits within the OSI model.
Figure
18: Where AppleTalk protocols fit in the OSI model
Like Novell's native
protocols, Apple's standard protocols are important because of Apple's
wide acceptance in the microcomputer market.
Layer-Four Standards:
Transport
The ISO has issued
a transport-layer standard that is simply called the Transport Protocol
(TP). Because it is an ISO standard, it is of worldwide importance.
At the transport layer,
Novell's native protocol is SPX™. SPX provides guaranteed packet delivery
and packet sequencing. Although it is basically a transport-layer protocol,
it also includes session-layer functions. The NetWare Core Protocol (NCP)
and SAP also provide transport-layer functions. SPX, NCP, and SAP will
be treated in more detail in a later section.
The AppleTalk protocol
set has a number of protocols that operate at the transport layer, including
Routing Table Maintenance Protocol, AppleTalk Echo Protocol, AppleTalk
Transaction Protocol, and the Name Binding Protocol.
IBM's NetBIOS protocol
(not shown in Figure 17) is also an important protocol at this layer and
at the session layer above.
The DOD's Transmission
Control Protocol, which is part of the TCP/IP standard, is important at
the transport layer to the same degree (extremely important) and for the
same reasons as the IP standard at layer three. This protocol provides
all functions required for this layer (transport) and part of the functions
for the session layer above.
Layer-Five Standards:
Session
The ISO session standard,
named simply "session," has the same worldwide importance as the ISO transport
standard. The DOD's Transmission Control Protocol performs important functions
at this layer.
In a NetWare environment,
the NetWare Core Protocol™ provides most of the necessary session-layer
functions. SAP also provides functions at this layer.
Layer-Six and Layer-Seven
Standards: Presentation and Application
Two important OSI protocols
encompassing both the presentation and application layers are File Transfer,
Access, and Management (FTAM) and Virtual Terminal Protocol (VTP). Each
of these protocols is exactly what its name implies. FTAM provides user
applications with useful file transfer and management functions. VTP supports
applications by converting specific terminal characteristics to a general
(virtual) terminal model shared by applications.
X.400 is an important
CCITT standard that encompasses both the presentation and application layers.
X.400 provides message handling and E-mail services. It is an important
standard because it is the basis for a number of pervasive E-mail packages
as well as for other widely used messaging products.
An important DOD standard
at this level is File Transfer Protocol (FTP), which, again, is named for
the service it provides.
The NetWare protocols
that provide presentation- and application-layer functions are NCP™ and
SAP. All NetWare protocols will be treated in more detail in a later section.
Further Perspective: Standards
and Open Systems
Product vendors' actual
implementation of OSI layers is even less neatly divided. Vendors implement
accepted standards, which already include mixed services from multiple
layers, in different ways.
So why go to all the
trouble to agree on a model and then define standards if you are not going
to be exact when fitting the standards to the model or in implementing
the standards when building a product?
Actually, standards
development and implementation have proceeded more or less as expected.
The OSI model was never intended to foster a rigid, unbreakable set of
rules. It was expected that in implementing the OSI communication model,
networking vendors would be free to use whichever standard for each layer
they deemed most appropriate. They would also be free to implement each
standard in the manner best suited for the purposes of their products.
As noted earlier, however,
it is clearly in a vendor's best interest to manufacture products that
conform to the intentions behind the OSI model. To do this, a vendor must
provide the services required at each OSI model layer in a manner that
will enable its system to be simply and easily connected to the systems
of other vendors—in other words, vendors must develop open systems. The
consequences of not doing so are severe and unavoidable.
Which leads to the
next issue—how do you determine if a system is an open system? You can
start by getting answers to simple questions such as: (1) Can you establish
communications using virtually any accepted communication standard? and
(2) How easily can you do this? For example, can you communicate with other
networks that are using the TCP/IP protocol, even if your network uses
some other protocol at that layer? If you can communicate, what kind of
effort is required? And how reliable are such communications?
As you begin asking
questions like these, you will find that Novell has the answers you need.
NetWare products support every standard we have presented, as well as virtually
every other accepted standard. The more you understand NetWare products,
the more you will understand that no system is more open than a NetWare
system.
The Network Operating
System
The network operating
system software acts as the command center, enabling all of the network
hardware and all other network software to function together as one cohesive,
organized system. In other words, the network operating system is the very
heart of the network.
Client-Server Network
Operating Systems
A client-server operating
system is responsible for coordinating the use of all resources and services
available from the server on which it is running.
The client part of
a client-server network is any other network device or process that makes
requests to use server resources and services. For example, network users
at workstations request the use of services and resources though client
software, which runs in the workstation and talks to the operating system
in the server by means of a common protocol.
On a NetWare client-server
network, users "log in" to the network server from the workstation. To
log in, a user enters a login command and gives his or her user name and
password. If the user name and password are valid, the server "authenticates"
the user and allows him or her access to all network services and resources
to which he or she has been granted rights. As long as the user has proper
network rights, the client-server operating system provides the services
or resources requested by the distributed applications running in workstations.
The operating system
manages various server resources, which include hardware such as hard disks,
RAM, printers, and equipment used for remote communications, such as modems.
The network file system is also a server resource.
In addition, the network
operating system provides many services, including coordinating file access
and file sharing (including file and record locking), managing server memory,
managing data security, scheduling tasks for processing, coordinating printer
access, and managing internetwork communications.
Among the most important
functions performed by a client-server operating system are ensuring the
reliability of data stored on the server and managing server security.
There are many other
functions that can and should be performed by a network operating system.
We do not have room to cover them all here. However, many functions might
be very important to you, and this means that choosing the right NOS is
of paramount importance. NetWare NOSs are robust systems that provide many
capabilities not found in less mature systems. NetWare NOSs also provide
a level of performance and reliability far above that found in most other
network operating systems.
To learn more about
client-server operating systems, including the services they can and should
provide, read the product sections that cover the IntranetWare™ and NetWare
client-server operating systems, including IntranetWare, IntranetWare for
Small Business, NetWare 4.11, NetWare 4.11 for OS/2, the NetWare 3.12 NOS,
and SFT III™ for IntranetWare.
Peer-to-Peer Network Operating
Systems
Peer-to-peer operating
systems have both advantages and disadvantages when compared to client-server
operating systems. They provide many of the same resources and services
as do client-server operating systems, and, under the right circumstances,
can provide good performance. They are also easy to install and are usually
inexpensive.
However, peer-to-peer
networks provide fewer services than client-server operating systems. Also,
the services they provide are a great deal less robust than those provided
by mature, full-featured client-server operating systems, and the performance
of peer-to-peer networks commonly decreases significantly under a heavy
load. Furthermore, except in the case of Novell's Personal NetWare™ peer-to-peer
network operating system, maintenance is often more difficult: Because
there is no method of centralized management, there are often many servers
to manage (rather than one centralized server), and many people may have
access to and the ability to change the configuration of different server
computers.
For more information
about the differences between peer-to-peer and client-server networks and
the level of services they offer, refer to the Personal NetWare product
description.
Desktop
Operating Systems
There are a number
of commonly used desktop operating systems, including Windows 3.x, Windows
NT, Windows 95, UNIX, PC-DOS, OS/2, MS-DOS, and various versions of the
Macintosh operating system.
Each of the different
desktop operating systems has advantages and disadvantages. Unfortunately,
for the most part, they are not compatible with each other. Software written
for one operating system will not function on another. Furthermore, peripheral
hardware (such as modems, facsimile machines, and so on) that is compatible
with the hardware required to run one kind of desktop operating system
is usually not compatible with hardware required to run other desktop operating
systems.
This brings us to another
important function of a network operating system—it should be able to interconnect
all of the commonly used desktop operating systems to ensure that all network
users have access to the computer that they are most familiar with and
that is best suited to the job they need to do.
Novell network operating
systems enable you to integrate all popular desktop operating systems directly
on one network. They allow this because they are able to translate the
data from one desktop operating system into data that the other desktop
operating systems can read.
Application
Software
One extremely important
issue to consider when selecting commercially built application software
is its degree of network and intranetwork integration. To effectively use
network and intranet services, application software must be well integrated
with the network operating system. The degree of network integration will
determine how well the application enables collaboration among network
users, whether and how well it provides direct access to all network services,
and whether it is as easy as it can be to manage across the network.
To learn more about
network-integrated software, see the product sections that cover Novell
GroupWare, including GroupWise™ 5.1, GroupWise WebAccess, GroupWise PhoneAccess,
and GroupWise Remote.
Internetworking
Why might a business
need subnetworks?
The most common reason
for segmenting a network is to preserve excellent network performance.
On even the fastest and most efficient network, if the network has too
many users (devices that need to transmit), the transmission media can
become so busy that devices have to wait an unacceptable time to transmit.
When this happens, users begin to notice delays when they try to save or
open files or perform other operations.
When you segment a
network, you give each subnetwork its own network address. This results
in two separate transmission media segments, which can be used simultaneously.
Each of the two segments will have only half the users of the original
network. Thus, you double network performance (on some networks, performance
can more than double because on an overloaded network, the overhead required
to manage transmission collisions takes a much larger percentage of bandwidth
than on a modestly busy network).
Networks are also segmented
to enhance data security and to minimize the effect of equipment failure
on any part of the network.
Internetworking includes
everything from connecting two small workgroup networks, each with perhaps
two or three workstations, to connecting thousands of computers—from notebook
computers to mainframes—on tens to hundreds of individual segments in a
worldwide organization.
Internetworking Devices:
Bridges and Routers
Software-based routers
and bridges can be part of a server's operating system or can at least
run in the server with the operating system. Software-based bridges and
routers can also be installed on standard computers to create dedicated,
standalone devices. For example, IntranetWare MultiProtocol Router software
is a family of software-based routing products that can be installed on
an IntranetWare, NetWare 4, or NetWare 3™ server or on a standalone PC.
To understand internetworking,
it is not essential that you understand all the technical differences between
a bridge and router. In fact, without some study, this can be a confusing
area. For example, if you read about IntranetWare MultiProtocol Routers,
you will find that these routers also perform what is called source-route
bridging.
However, without a
basic understanding of bridging and routing technology (and related terminology),
you will find it difficult to understand the capabilities of some products
and the reasons such capabilities are useful or important. Please keep
in mind throughout the following discussion that bridges and routers have
one important thing in common: They both allow the transfer of data packets
(frames) between subnetworks with different network addresses.
Bridges
Simple bridges are
used to connect networks that use the same physical-layer protocol and
the same MAC and logical link protocols (OSI layers one and two). Simple
bridges are not capable of translating between different protocols.
Other types of bridges,
such as translational bridges, can connect networks that use different
layer-one and MAC-level protocols; they are capable of translating, then
relaying, frames.
After a physical connection
is made (at OSI layer one), a bridge receives all frames from each of the
subnetworks it connects and checks the network address of each received
frame. The network address is contained in the MAC header. When a bridge
receives a frame from one subnetwork that is addressed to a workstation
on another subnetwork, it passes the frame to the intended subnetwork.
Figure 19 illustrates, in a general fashion, how a bridge relays frames
between subnetworks.
Figure
19: Internetworking through a bridge
A bridge assumes that
all communication protocols used above the data-link layer at which it
operates (OSI layers three through seven) are the same on both sides of
the communication link. Of course, this must be true, or there must be
translation between unlike protocols at layers three through seven for
the receiving computer to be able to interpret the transferred data.
Spanning Trees and Source-Route
Bridging
Spanning trees prevent
problems resulting from the interconnection of multiple networks by means
of parallel transmission paths. In various bridging circumstances, it is
possible to have multiple transmission routes between computers on different
networks. If multiple transmission routes exist, unless there is an efficient
method for specifying only one route, it is possible to have an endless
duplication and expansion of routing errors that will saturate the network
with useless transmissions, quickly disabling it. Spanning trees are the
method used to specify one, and only one, transmission route.
Source-route bridging
is a means of determining the path used to transfer data from one workstation
to another. Workstations that use source routing participate in route discovery
and specify the route to be used for each transmitted packet. Source-route
bridges merely carry out the routing instructions placed into each data
packet when the packet is assembled by the sending workstation—hence the
name "source routing." In discussions of bridging and routing, do not be
confused by the term "source routing." Though it includes the term "routing,"
it is a part of bridging technology. Source-route bridging is important
because it is a bridge-routing method used on IBM Token-Ring networks.
You should understand
that bridging technologies and routing methods can be combined in various
ways. For example, there is an IEEE specification for a source-route transparent
bridge, a bridging scheme that merges source-route bridging and transparent
bridging in one device.
From this simple discussion
of bridging, one thing should be apparent: When choosing internetworking
products, it is important to select those that support the various bridging
methods—products such as IntranetWare MultiProtocol Router. (For further
details, see the IntranetWare MultiProtocol Router 3.1 product section.)
Routers
Like some bridges,
routers can allow the transfer of data between networks that use different
protocols at OSI layers one and two (the physical layer and the data-link
layer, which includes sublayers for media access control and logical link
control). Routers can receive, reformat, and retransmit data packets assembled
by different layer-one and layer-two protocols. Different routers are built
to manage different protocol sets. Figure 20 illustrates how a router transfers
data packets.
Figure
20: Internetworking through a router
NetWare Internetworking
Protocols
Figure
21: Where NetWare protocols fit in the OSI model
Each of the native
NetWare protocols shown in Figure 21 plays a role in NetWare internetworking,
either directly or indirectly.
IPX: The Network Layer
Protocol
In a NetWare environment,
internetwork packet routing is accomplished at the network layer. Thus,
IPX is the NetWare protocol that addresses and routes packets between internetworked
computers.
IPX bases its routing
decisions on the address fields in its packet header (provided by the MAC
protocol) and on the information it receives from other NetWare protocols.
For example, IPX uses information supplied by either the RIP or NLSP™ protocols
to forward packets to the destination computer or to the next router. IPX
also uses SAP.
RIP and NLSP: The Routing
Protocols
RIP: A Distance-Vector
Routing Protocol
In an internetwork
using distance-vector routing, routers periodically determine if the internetwork
configuration has changed. They also periodically broadcast packets to
their immediate neighbors; these packets contain all information they currently
have about the internetwork's topology.
After receiving any
information, distance-vector routers consolidate the information and pass
summarized data along to other routers, servers, and end devices, such
as printers and workstations. Through this periodic checking and broadcasting,
which is performed at regular intervals regardless of whether the internetwork
has changed, all routers are kept updated with correct internetwork addresses
for all computers and other connected devices, as well as with the best
route for transferring data between any two devices.
Because RIP is a distance-vector
protocol, NetWare routers that use RIP work in the way described above,
performing periodic checking and information exchange and updating their
routing tables with any new information.
RIP is one of a number
of well-known distance-vector routing protocols. Examples of other such
protocols include IP RIP and Cisco IGRP, part of the IP protocol suite,
and RTMP, part of the AppleTalk protocol suite.
NLSP: A Link-State Routing
Protocol
Link-state protocols,
a relatively recent development, adapt more quickly to network topology
changes than do distance-vector protocols. Thus, they are better than distance-vector
protocols for managing internetworking on large, complex internetworks.
In an internetwork
that uses a link-state routing protocol, each router or server provides
information about itself and its immediate neighbors to every reachable
router in a routing area. Each router's map includes all the area's routers
and servers, the links connecting them, and the operational status of each
router and link. However, each router builds its own routing map rather
than relying on secondhand summaries, as do distance-vector routers. Also,
routing transmissions are made only when the internetwork changes, not
at predefined intervals. Thus, networks using link-state routing are not
burdened by unnecessary routing traffic.
Because NLSP works
as explained above, it significantly reduces the communication overhead
required for routing. NLSP can significantly improve network performance
because it frees resources to be used for transferring data packets rather
than routing information. NLSP is particularly efficient for wide area
network routing, where available communication bandwidth is ordinarily
limited.
Examples of other link-state
protocols include the Open Shortest Path First protocol, part of the TCP/IP
protocol suite, and the Intermediate System-to-Intermediate System protocol,
a router-to-router protocol that is part of the OSI suite.
As a matter of note,
various link-state and distance-vector routing protocols can coexist on
the same NetWare internetwork and even in the same IntranetWare MultiProtocol
Router. Furthermore, individual routers can be configured to accept or
to reject individual protocols.
SAP: Service Advertising
Protocol
Servers and routers
use SAP to advertise their services and network addresses. SAP enables
network devices to constantly correct their information about which network
services are available. While servers are running, they use SAP to inform
the rest of the network of the services they offer. When a server goes
down, it uses SAP to inform the network that its services are no longer
available.
Routers gather service
information and share it with other routers. Workstations use the information
made available through SAP to obtain the network addresses of servers that
offer the services they need.
NCP: NetWare Core Protocol
NCP does not play a
direct role in routing. However, it does provide session control and packet-level
error checking between NetWare workstations and routers.
SPX: Sequenced Packet
Exchange
Like NCP, SPX does
not play a direct role in routing. SPX is connected with internetworking
only in that it guarantees delivery of all routed packets.
Gateways
A gateway may connect
dissimilar systems on the same network or on different networks (thus,
using a gateway does not necessarily involve internetworking). For example,
a gateway might translate protocols at several different OSI layers to
allow transparent communications between NetWare IPX-based systems and
systems based on TCP/IP, System Network Architecture (SNA), or AppleTalk.
Figure 22 illustrates how a gateway is used to translate protocols to enable
communications between two heterogeneous systems.
Figure
22: Gateways provide protocol translation between dissimilar systems at
more than one OSI layer.
A gateway may consist
of hardware, software, or a combination of the two, and it may provide
translation at all or at only some of the different OSI layers, depending
on the types of systems it connects.
There are a number
of NetWare gateways that provide access to computer systems not based on
the native NetWare/IPX protocol suite. NetWare for Macintosh is a software-based
gateway that connects Macintosh computers to a PC-server-based NetWare
network. NetWare for SAA is a gateway that enables NetWare users to transparently
access SNA-based IBM hosts.
Real
World Networking
One Server, Multiple Networks
of the Same Type
For example, one server
could contain two Ethernet network adapters, each supporting a different
cable segment. There could be several computers connected to each cable
segment, in a star physical layout, with each cable segment using contention
(CSMA/CD) for the media access control. Each of the cable segments would
have a different network address—thus, each would be an independent subnetwork.
Together, the two separate
networks would form an internetwork, connected by means of internal routing
capabilities built into the server. (Remember, we have already said that
in NetWare servers, internetworking is accomplished through routing at
the network layer.)
Figure 23 illustrates
the one-server internetwork described above.
Figure
23: Internetworking two networks using the same type of network adapter
(MAC) in one NetWare server, by means of the server's internal routers
In the case of the
above network, routing would be accomplished using the NetWare IPX™ protocol
or the NetWare IP protocol, with support from the other NetWare routing
protocols, as previously described.
Every NetWare server
is capable of using internal routers to accomplish local network routing
by means of the NetWare routing protocol set and AppleTalk. All NetWare
internal routers operate at layer three of the OSI model and are for use
with small workgroup or departmental networks. For larger or more complicated
internetworks, or for departments with heavy server-processing requirements,
the IntranetWare MultiProtocol Routers or dedicated routers from other
vendors provide the necessary extra routing power and capabilities.
In a slightly more
complex internetwork, a NetWare server could support multiple cable segments
using the same physical layouts but different media access controls.
For example, a server
could contain one Ethernet network adapter and one token-ring network adapter,
with a cable segment attached to each. The Ethernet network might be connected
in a physical star and use CSMA/CD for the media access control. The token-ring
network might also be connected in a physical star, but it would use token
passing for media access control. Like the simpler configuration explained
in the previous section, each cable segment would have a different network
address. Figure 24 illustrates this more complex one-server internetwork.
Figure
24: Internetworking two networks using different types of network adapters
(MAC) in one NetWare server, by means of the server's internal routers
In the case of the
internetwork shown above, routing would again be accomplished using the
NetWare IPX or NetWare IP protocol, with support from the other NetWare
routing protocols.
The two one-server
networks we have seen each support only two separate subnetworks. As a
matter of note, all NetWare servers are capable of supporting as many as
four different network adapters (four separate subnetworks), in any combination
of same or different types.
Please notice that
even though the token-ring network above was described as a physical star,
it is drawn as a ring to signify that it is a token-ring network (which
uses token passing as the media access control). We will adhere to this
convention throughout this primer because in virtually all illustrations,
it will be more important to make the logical topology clear than to be
concerned with the physical topology.
Multiple Servers With
Multiple Networks of Different Types
For example, a complex
internetwork might consist of two one-server subnetworks connected by a
standalone router, such as the IntranetWare MultiProtocol Router. Each
server might contain multiple network interface adapters.
One server might contain
two Ethernet network adapters and one token-ring network adapter, with
a cable segment attached to each. One of the Ethernet adapters might support
a PC network, and the other Ethernet adapter might connect to both PCs
and Macintosh computers. The NetWare for Macintosh product running on the
server would support the Macintosh computers.
The other server might
contain one Ethernet adapter and one ARCnet adapter, with the Ethernet
adapter again supporting both PCs and Macintoshes, and the ARCnet adapter
supporting a cable segment with a number of PCs attached.
Each of the two servers
would have a unique internal number (server address), and each cable segment
in each server would have a unique physical network (cable segment) address.
In this case, there
would be five subnetworks on the internetwork, three attached to one server
and two attached to the other. The internal server routers would accomplish
the routing between any two workstations on subnetworks attached directly
to the same server. Both the internal server routers and the intermediate
standalone router would be involved in the routing between any two workstations
on subnetworks attached to different servers.
Figure 25 illustrates
the two-server internetwork described above.
Figure
25: Internetworking multiple networks using different types of network
adapters (MAC) in two NetWare servers, by means of internal and standalone
routers
Host
Access
Host systems can provide
access to additional application software, additional resources such as
data storage devices and printers, and additional processing power. For
example, you might want to log in to an IBM AS/400 minicomputer to run
an application available only on that computer or to use its processing
power for one task while you were using the processing power of your own
workstation for some other task. Or, you might want to print a large report
on a high-speed printer connected to the AS/400.
The illustration in
Figure 26 shows a multiserver NetWare network with an IBM mainframe, an
IBM AS/400 minicomputer, and several UNIX workstations connected as host
computers.
Figure
26: Host systems connected to a complex multiserver NetWare network
A number of leading
networking companies have entered into original equipment manufacturer
(OEM) partnerships with Novell. Many provide NetWare connectivity to host-based
environments.
Wide
Area Networking
The traditional definition
of wide area networking has been "connecting two or more networks existing
at widely separate geographic sites." Some traditionalists also prescribe
that the separate networks must be connected by means of common carrier
telecommunication facilities (private companies that rent resources such
as T1 lines and microwave transmission equipment). For the purposes of
this primer, we'll use the first, general definition and let you decide
how to apply it in specific internetworking cases. But, to give you some
background to help you make such decisions, let's discuss a few specific
internetworking cases and a few terms related to wide area networking.
Of course, like any
general term used in connection with rapidly changing technology, not everyone
will agree on an exact definition of wide area networking. What is "widely
separate"? And, does the connection really have to be through a common
carrier? Many major companies now own their own equipment linking networks
many miles apart.
Let's look at some
examples. Suppose you connect two networks in two different buildings 100
yards apart by means of asynchronous modems and common telephone lines.
Is that wide area networking? Most knowledgeable computer networking people
would say no—this would be "one-site" or "campus" networking. What if the
networks were two miles apart and separated by a major interstate highway?
Or, what if they were 15 miles apart, on opposite sides of a major city?
There are many computer networking people who would still not call this
wide area networking; they might use a recently coined term—"metropolitan
area networking." Others consider metropolitan area networking a part of
wide area networking. Of course, everyone would agree that two networks
connected on opposite sides of a continent by means of a satellite microwave
link rented from a common carrier is an example of a wide area network.
You can decide for
yourself where you think wide area networking begins and ends. Now let's
look at some general possibilities.
Figure 27 shows two
separate branch office internetworks connected to a third internetwork
at a main corporate office. Each of the three existing internetworks has
multiple servers and existing host connections. One of the branch office
networks is connected to the corporate network by means of asynchronous
modems and regular voice-grade telephone lines. The other branch office
network is connected by means of a common carrier-provided intermediate
link—in this case an X.25 packet-switching network. Examples of such networks
include Tymnet or Telnet.
However, either network
could be connected by other means that we have discussed, such as frame
relay or a dedicated leased line link, perhaps using PPP.
The following section
describes important WAN and LAN technologies in greater detail.
Figure
27: Wide area networking: three networks at widely separated sites connected
through asynchronous modems and an X.25 connection
Important WAN and High-Speed
Technologies
-
100Base-T
-
100VG-AnyLAN
-
Fiber
Distributed Data Interface (FDDI)
-
X.25
-
Frame
relay
-
Asynchronous
Transfer Mode (ATM)
-
Integrated
Services Digital Network (ISDN)
-
Synchronous
Optical Network (SONET)
100Base-T
Distinguishing Characteristics
CSMA/CD was known to
be scalable before the 100Base-T standard was created. A scaled-down version
of Ethernet (1Base-5) uses CSMA/CD, provides data transfer rates of 1 Mbit/s,
and enables longer transmission distances between repeaters. If CSMA/CD
could be scaled down, then it could be scaled up. Specifying changes such
as decreased transmission distances between repeaters produced a reliable
data transfer rate of 100 Mbit/s, 10 times faster than traditional 10Base-T
Ethernet.
100Base-T supports
Category 3 and 5 unshielded twisted-pair (UTP) wiring, Type-1 shielded
twisted-pair (STP) wiring, and fiber-optic cable. It uses four wire pairs
of Category 3 UTP cable—three for data and one for collision detection.
However, 100Base-T uses only two wire pairs of Category 5 UTP cable.
Figure
28: On 100Base-T networks, the physical topology is a star and the logical
topology is a bus. A broadcast signal travels to all parts of the cable.
Advantages
In addition, it's easy
to upgrade from 10Base-T Ethernet to 100Base-T Ethernet. Both traditional
10Base-T and 100Base-T Ethernet use CSMA/CD, and some network cards now
support both 10 Mbit/s and 100 Mbit/s Ethernet. The adapter cards automatically
sense whether it is a 10 Mbit/s or 100 Mbit/s environment and adjust their
speed accordingly. Because 100Base-T and 10Base-T Ethernet can coexist,
network supervisors can upgrade network stations from 10Base-T to 100Base-T
one at a time, as needed. Also, most network supervisors are already familiar
with CSMA/CD, so there is no need for expensive retraining.
100Base-T can be an
inexpensive way to make your network faster. Adapter cards are not significantly
more expensive than 10Base-T cards. In addition, Category 3 and Category
5 UTP cable are relatively inexpensive and many organizations already have
either Category 3 or 5 cable installed.
Disadvantages
In addition, the fact
that 100Base-T is based on CSMA/CD creates problems. 100Base-T may scale
CSMA/CD to its limit, making 100 Mbit/s the maximum data transfer rate
for this standard. To increase data transfer rates, 100Base-T specifies
shorter distances between signal repeaters, and these distances may be
as short as is practical. Also, because CSMA/CD is a shared media contention
scheme, collisions will occur, especially under maximum loads. This results
in increased overhead, which reduces actual data throughput.
Furthermore, 100Base-T
requires four wire pairs of Category 3 cable, and not all companies that
have Category 3 cable have four wire pairs available. Thus, companies that
are already using some wire pairs for a different purpose, or that installed
cable with fewer than four wire pairs or cable that does not meet Category
3 standards, will have to recable to use 100Base-T.
100VG-AnyLAN
Distinguishing Characteristics
For example, when a
workstation needs to transmit, it signals the intelligent hub that it needs
access to the transmission media. If the intelligent hub receives several
requests, it will give access to the workstation that has the highest priority.
(100VG-AnyLAN will also function without a prioritization scheme.) If the
workstations requesting access have the same priority, the intelligent
hub will assign the token to the workstations in the order they request
access to the transmission media.
In addition, 100VG-AnyLAN
supports both Ethernet and token-ring networks. It also supports Category
3 and 5 UTP, Type-1 STP, and fiber-optic cable. 100VG-AnyLAN uses four
wire pairs of Category 3 or Category 5 UTP cable.
Figure
29: On 100VG-AnyLAN networks, both the physical and logical topologies
are stars. The signal from one node goes to the intelligent hub and is
routed only to the correct destination node.
Advantages
Furthermore, unlike
100Base-T, 100VG-AnyLAN supports token-ring networks as well as Ethernet,
providing data transfer rates as high as 100 Mbit/s to the former.
100VG-AnyLAN and 100Base-T
also share many advantages. The cost of 100VG-AnyLAN is comparable to 100Base-T:
Adapter cards that support both 10 and 100 Mbit/s are not priced significantly
higher than traditional 10Base-T Ethernet cards. Both standards also support
the same types of transmission media. In addition, both are easy to upgrade.
Disadvantages
Previously installed
cable may be problematic for 100VG-AnyLAN networks, as it is for 100Base-T.
100VG-AnyLAN uses all four wire pairs of Category 3 or 5 UTP cable. Thus,
companies that are already using some wire pairs for a different purpose,
or that installed cable with less than four wire pairs or cable that does
not meet Category 3 standards, will have to recable to use 100VG-AnyLAN.
Fiber
Distributed Data Interface
FDDI is officially
designated as ANSI X3T9.5 and operates at the physical and data-link layers
(levels one and two) of the OSI model. Like 100Base-T and 100VG-AnyLAN,
FDDI provides data transfer rates as high as 100 Mbit/s.
Figure
30: A simple server-based backbone connecting two LAN segments
Distinguishing Characteristics
On FDDI networks, every
node acts as a repeater. FDDI supports four kinds of nodes: dual-attached
stations (DASs), single-attached stations (SASs), single-attached concentrators
(SACs), and dual-attached concentrators (DACs). DASs and DACs attach to
both rings; SASs and SACs attach only to the primary ring. Several SASs
often attach to the primary ring through a concentrator so that an SAS
failure will not bring down the entire network. If the cable is cut or
a link between nodes fails, DASs or DACs on either side of the failure
route signals around the failed segment using the secondary ring to keep
the network functioning.
FDDI uses token passing
for the media access control method and is implemented using fiber-optic
cable.
Figure
31: If a cable section on an FDDI network goes down, DASs on either side
of the failed section automatically reconnect the primary and secondary
rings. Also note that the server has a redundant connection to improve
reliability.
Advantages
Disadvantages
X.25
Distinguishing Characteristics
Figure
32: X.25 networks are often provided by telecommunication carriers. CompuServe
uses X.25 on its network.
Advantages
Disadvantages
Frame
Relay
Distinguishing Characteristics
Frame relay services
are typically provided by telecommunications carriers. Customers install
a router and lease a line (often a T1 or fractional T1 line) to provide
a permanent connection from the customer's site to the telecommunications
carrier's network. This connection enables frame relay to use permanent
virtual circuits (PVCs), which are predefined network paths between two
locations.
With frame relay, the
router encapsulates (or frames) network layer packets, such as IP and IPX
packets, directly into a data-link level protocol and sends them on to
the packet-switched network. Like X.25, frame relay uses variable-size
frames, but it eliminates the error checking required on X.25 networks.
A frame relay switch simply reads the header and forwards the packet, perhaps
without even fully receiving a frame before forwarding it. Intelligent
end stations must identify missing or corrupted frames and request retransmission.
Figure
33: Frame relay is a WAN technology that enables companies to connect LANs
through a telecommunicationscarrier's network. AT&T WorldNet Intranet
Connect Service currently uses this technology.
Advantages
Although frame relay
is fairly complex to implement, value-added resellers and some telephone
companies will assist customers in determining their needs and will help
install the technology.
Disadvantages
In addition, frame
relay is more complex to implement than X.25. Customers must negotiate
a service agreement with the phone company, lease a line, and have it installed.
They must also purchase and install a frame relay-compatible router.
Asynchronous
Transfer Mode
ATM is extremely scalable;
data transfer rates range from 25 Mbit/s to 2.4 gigabits per second (Gbit/s).
This wide range of data transfer rates reflects the various ways in which
ATM can be used. The 25 Mbit/s rate is a new offering meant for desktop
environments. In LAN backbones, ATM provides data transfer rates of 100
Mbit/s and 155 Mbit/s. At the high end, WAN implementations using ATM and
SONET together have achieved data transfer rates of 2.4 Gbit/s. (For more
information about SONET, see the "Synchronous Optical Network" heading
later in this primer.)
Distinguishing Characteristics
In a LAN implementation,
ATM functions at the data-link layer's media access control sublayer. It
further divides the MAC sublayer into three layers: LAN Emulation, ATM
Adaptation Layer (AAL), and ATM. LAN Emulation enables you to integrate
ATM with Ethernet and token-ring networks without modifying existing Ethernet
or token-ring protocols.
On a mixed network,
LAN Emulation hardware sits between the Ethernet or token-ring segment
and the ATM part of the network. It uses the three layers mentioned above
to convert packets moving toward the ATM segment into cells and to assemble
cells moving toward the Ethernet or token-ring segment into packets. AAL
and ATM put data into standard-sized cells. In most network computing situations,
ATM Adaptation Layer 5 breaks packets into 48-byte blocks that are then
passed to the ATM layer, where the five-byte header is attached to form
a complete 53-byte cell.
Advantages
One reason that ATM
is so fast is its use of cells. Because cells are a standard size, ATM
networks handle data in a predictable, efficient manner at the switches.
Standard-sized cells and high-bandwidth media like fiber-optic cable also
enable ATM to support real-time voice, video, and data traffic.
ATM also offers flexibility
in its transmission media. As many as 22 ATM specifications exist for media
like unshielded twisted-pair, shielded twisted-pair, and fiber-optic cable.
(ATM is generally implemented with fiber-optic cable.)
Although it is seen
as a technology of the future, ATM can currently be integrated with Ethernet
and token-ring networks, through use of LAN Emulation.
Disadvantages
Integrated
Services Digital Network
Distinguishing Characteristics
ISDN offers Basic Rate
Interface (BRI) for individuals or small branch offices and Primary Rate
Interface (PRI) for larger companies.
BRI uses two bearer,
or B, channels (providing 64 kbit/s each) to transmit and receive data
and one delta, or D, channel for call setup and management.
PRI is the same thing
as a T1 line. A T1 line in the United States consists of 23 B channels
and one D channel, providing a total data transfer rate of 1.544 Mbit/s.
A T1 line in Europe consists of 30 B channels and one D channel, providing
a total data transfer rate of 2.048 Mbit/s. A fractional T1 uses only some
of the B channels in a T1 line (and thus offers some fraction of the total
T1 data transfer rate).
ISDN requires special
equipment at the customer's site, including a digital phone line and a
network termination unit (NT-1). An NT-1 converts the bandwidth coming
over the line into the B and D channels and helps the phone company with
diagnostic testing. The NT-1 also provides a connection for terminal equipment,
such as ISDN telephones and computers that have an ISDN interface. In addition,
the NT-1 provides terminal adapter (TA) equipment to connect equipment
that is not compatible with ISDN. TA equipment provides an intermediary
connection point: Such equipment has an ISDN interface, for connection
to the NT-1, and a non-ISDN interface, for connection to non-ISDN equipment.
Advantages
With ISDN, you can
transmit voice and data traffic simultaneously: An ISDN user can simultaneously
talk on the phone and download a data file to his or her computer, over
the same telephone line. For example, one BRI ISDN configuration enables
users to use the two B channels (128 kbit/s) for data and part of the D
channel for a phone conversation.
Disadvantages
Acceptance of ISDN
in the United States has been slow for several reasons. First, to understand
ISDN well enough to even order services requires considerable effort. Furthermore,
configuration can be difficult. In addition, ISDN lacks the standards that
ensure interoperability. As a result, customers must be careful to purchase
equipment that is compatible with the local phone company's equipment.
Another problem is that not all phone companies offer the same services,
so customers must ensure that the services they need are available in their
area. Finally, to take full advantage of ISDN, customers must communicate
with others who also have ISDN.
Synchronous
Optical Network
Distinguishing Characteristics
Data communications
sometimes prove difficult because digital signaling rates can vary. For
example, in the United States, a T1 line provides 1.544 Mbit/s; in Europe,
a T1 line (sometimes called an E1 line) provides 2.048 Mbit/s. SONET resolves
such problems by defining how switches and multiplexers coordinate communications
over lines with different speeds, including defining data transfer rates
and frame format.
SONET defines a number
of Optical Carrier (OC) levels. Each level defines an optical signal and
a corresponding electrical signal called Synchronous Transport Signal (STS).
The base level is OC-1/STS-1 or 51.84 Mbit/s. Each level's rate is a multiple
of 51.84 Mbit/s. The table below shows the OC levels and the corresponding
data transfer rates that SONET defines.
|
OC
Level
|
Data
Rate
|
|
OC-1
|
51.8
Mbit/s
|
|
OC-3
|
155.5
Mbit/s
|
|
OC-9
|
466.5
Mbit/s
|
|
OC-12
|
622.0
Mbit/s
|
|
OC-18
|
933.1
Mbit/s
|
|
OC-24
|
1.24
Gbit/s
|
|
OC-36
|
1.86
Gbit/s
|
|
OC-48
|
2.48
Gbit/s
|
SONET also provides
easy access for low-speed signals, such as DS-0 (64 kbit/s) and DS-1 (1.544
Mbit/s) by assigning them to sub-STS-1 signals called Virtual Tributaries.
Advantages
Disadvantages
Global
Networking
The global networking
ideal is the simple, powerful idea of people around the world connecting
to a network on which they can share ideas, exchange information, and access
endless electronic resources. Novell will play a major role in making the
global networking ideal a reality. Although the ideal is still some years
away, a basic form of global networking, based on the Internet and the
intranet, exists today.
The Internet
The Internet is a global
network, but in many ways, it does not currently meet the global networking
ideal. From a business standpoint, for example, the Internet has several
disadvantages. First, the Internet uses packet-switching, so you can never
be sure what route a packet will take or how long it will take to arrive.
Second, because no one owns the Internet, no one is responsible for ensuring
that the network as a whole is functioning properly (or has the authority
to require that it is). Third, while improvements have been made, security
on the Internet is still a problem. Fourth, the Internet does not offer
the fastest data transfer rates available.
Intranets
The term "intranet"
sprang up virtually overnight when companies discovered that they could
use publicly available Internet technologies to make useful information
immediately available to all employees, no matter where the employees were
located; that they could still secure the information from unwanted access;
and that, along with these other advantages, they could also make the information
available at the lowest possible cost.
On a typical intranet,
there is a World Wide Web server, on which information is published in
an electronic format called Hypertext Markup Language (HTML). Workstations
have some type of client software, most often a Web browser, through which
they can access any information published (in HTML format) on any Web server.
Users of client stations can be given different rights so that they can
access only selected information on selected Web servers.
The main reason for
a company to implement an intranet is that an intranet enables a business
to collect, manage, and disseminate information more quickly and easily
than ever before, even much more quickly and inexpensively than with other
current means of electronic communications, including E-mail and other
types of cross-platform publishing. In fact, intranet publishing is the
ultimate in cross-platform publishing because it is based on the Internet
technologies that were developed specifically for the purpose of allowing
information sharing among dissimilar computing systems.
While even a small
company with only one office and a small network can benefit from an intranet,
the value of an intranet increases with the number of employees, the size
of the network, and the number of geographically separate sites. The reason
is that as a company grows, if the company continues to use conventional
means of information dissemination, such as printed memoranda and newsletters,
the cost of disseminating information to all employees increases exponentially.
And other methods of sharing information, such as E-mail and file sharing,
also fall short of the cost savings and immediacy that can be obtained
through intranet publishing.
On an intranet, any
employee with a properly configured workstation and a Web browser can read
documents as soon as the files are completed and copied to any Web server,
regardless of where the employee is located. If a company were to instead
disseminate documents as files in a public directory or by E-mail, the
documents would have to be provided in multiple formats to accommodate
the various computing platforms and applications used within the company.
There would need to be people dedicated to the task of preparing the differently
formatted documents and disseminating them to different locations where
they could be accessed. In even a small company, this type of effort takes
significantly more time and costs far more than does publishing the same
information once, in HTML format, on a single Web server. In a large company,
the time and cost differences can be enormous.
Intranet publishing
has other advantages. One important advantage is that the network can update
your intranet documents automatically, in real time. For example, if you
published a document that contained the stock price for your company or
news about the market in which your company competes, you could create
a Web server script that would automatically update the document every
15 minutes with the most current stock price and market news. With immediate
access to up-to-date information, employees can respond more quickly to
changes in the marketplace (with the result of increased profits). Also,
after the script is created, the network continues to update the information—the
work isn't forgotten or ignored because employees are too busy—and there
is no further cost.
In addition, you can
get immediate feedback about the documents published on your intranet.
For example, with paper-based documents or publicly available files stored
on a server, you cannot determine whether or not people are reading the
documents. If you published the documents on an intranet server, however,
the network could track how many people read the documents and which documents
were used the most.
Businesses are continually
finding more ways to use intranets to decrease costs, especially since
the specification for World Wide Web documents has been extended to include
graphics, audio clips, and movies. For example, many companies have installed
applications that allow employees to access company databases directly
from a Web browser, thus avoiding the cost of specialized database access
programs. Recent products such as Novell's GroupWise WebAccess even allow
employees to read their E-mail messages and schedules directly from a Web
browser.
Another factor that
makes any intranet valuable is that after it is built, it can be connected
to the Internet with very little extra effort. Remote users, such as traveling
employees, suppliers, and customers, can then access your intranet documents
over the Internet. You can control access to your intranet documents, allowing
the general public to view some documents and allowing only authorized
users to view others. Furthermore, you can allow employees to connect to
the Internet and access a vast pool of information that covers nearly every
topic imaginable.
Of course, intranets
need not be connected to the Internet: An intranet may be only local, or,
if it is a WAN intranet, the various locations might be connected by means
other than the Internet. However, many intranets are now connected to the
Internet, and in the future many more will be. The most important reason
is that the Internet is a ready-made, low-cost WAN backbone. And, as mentioned
above, if your intranet is connected to the Internet, all users can easily
and speedily access the wealth of information available there.
Extranets
On an extranet, each
connected company usually makes some selected part of its intranet accessible
to the employees of one or more of the other companies. For example, several
companies might create an extranet to consolidate data gathering and share
data, or to jointly develop and share training programs and other material,
or to coordinate project management for a common work project. On an extranet,
each company uses the security inherent in its own intranet to the keep
employees of other companies from accessing information they do not need
to see.
The collaborative business
application is a powerful extranet tool. Such applications, possibly developed
jointly by participating companies, enable the employees of the different
companies to work together very effectively without leaving their offices
(which might be located in different places all over the world).
For example, a consumer
company might work with a supply company to connect their intranets and
create a supply ordering system, to allow all employees of the consumer
company to order whatever supplies they needed, whenever they needed, directly
from the supply company. A consumer company employee might order by using
his or her Web browser to look through one or more electronic catalogs
that the supply company published on the extranet. The employee might check
a box next to each of the items he or she needed. Different employees might
be given different rights to different catalogs so that they could see
and order only from selected parts of a catalog. Also, different employees
might be allowed to see different items in each part. Underlying parts
of the collaborative business application could sort all ordered items
by company division, group, and employee and fill out one daily purchase
requisition containing all items ordered by all employees. Each purchase
requisition could be immediately delivered over the extranet. For the supply
company, the application could automatically generate a shipping ticket
that contained the items to be shipped, broken down by division, group,
and the person each item was to be delivered to.
For the consumer company,
the end result might be to eliminate the need to stock any supplies and
to considerably reduce purchasing costs. The consumer company employees
might be able to get any supplies they needed in less time than ever before.
And the supply company might be able to sell more supplies and deliver
them faster than before, with less staff than before.
Because almost all
intranets and extranets will eventually be connected to the Internet, intranet
technology should be designed to deal as effectively as possible with the
security problems and other problems inherent to the Internet. Thus, Novell
is constantly working on new technologies such as IntranetWare Border Services,
which you can read about in the Early Access Release section.
The Information Superhighway
The ideal global network,
or Information Superhighway, will include a vastly improved Internet and
many other networks, services, and technologies. The Information Superhighway
will be pervasive: Low-cost access will be available to virtually everyone
worldwide. The Information Superhighway will provide homes, businesses,
and other organizations with a myriad of services, such as on-demand video,
E-mail, electronic commerce, shopping, research, video conferencing, and
voting services. In sum, the Information Superhighway will provide literally
every digitally deliverable service to everyone on the globe.
Presently, the Information
Superhighway is only a concept, but governments, businesses, and public
institutions worldwide are taking steps to make the Information Superhighway
a reality.
Novell contributes
to the growth of the Information Superhighway with all of its offerings,
from core infrastructure and services such as the IntranetWare and NetWare
4.11 operating systems; to advanced services such as Novell Directory Services™,
Novell Connect Services™, NetWare Telephony Services™, and the Novell Web
Server 3.0; to Internet and intranet access products such as the LAN WorkPlace®
and LAN WorkGroup™ family of products; to advanced groupware and intranet
applications such as GroupWise 5.1 and GroupWise WebAccess.
We hope this primer
has been helpful to you. We welcome your comments and suggestions. Happy
networking!
Primer Appendix
ASCII Code Set
In the ASCII coding
scheme, information (a numeral, symbol, or alphabetic character) is represented
by the value of a data unit called a "byte." Each byte can represent one
character. There are eight bits in one byte. Bits, short for binary digits,
are the data units actually stored in the computer as either a one or a
zero. Computers read stored bytes and interpret them as the codes that
represent character-based information.
Sample Byte (Bit Settings)
8 7 6 5 4 3 2 1
1 1 0 0 0 0 0 1 = 65 = letter "A"
0 0 1 1 1 0 0 0 = 56 = number "8"
ASCII Code-to-Character
Conversion
