Upgrading and Repairing Networks 4th Edition Scott Mueller - The ebook in PDF and DOCX formats is ready for download
Upgrading and Repairing Networks 4th Edition Scott Mueller - The ebook in PDF and DOCX formats is ready for download
https://ebookultra.com/download/the-8051-microcontroller-4th-edition-
scott-mackenzie/
https://ebookultra.com/download/create-your-own-website-4th-edition-
scott-mitchell/
https://ebookultra.com/download/refurbishment-and-upgrading-of-
buildings-2nd-edition-christopher-a-gorse/
https://ebookultra.com/download/insall-scott-surgery-of-the-
knee-2-volume-set-4th-edition-w-norman-scott/
Relational Justice Repairing the Breach 1st Edition
Jonathan Burnside
https://ebookultra.com/download/relational-justice-repairing-the-
breach-1st-edition-jonathan-burnside/
https://ebookultra.com/download/building-cisco-multilayer-switched-
networks-bcmsn-4th-edition-edition-froom/
https://ebookultra.com/download/local-economic-and-employment-
development-skills-upgrading-new-policy-perspectives-oecd/
https://ebookultra.com/download/mastering-iis7-implementation-and-
administration-john-paul-mueller/
https://ebookultra.com/download/upgrading-to-microsoft-
office-2010-1st-edition-mary-terese-cozzola/
Upgrading and Repairing Networks 4th Edition Scott
Mueller Digital Instant Download
Author(s): Scott Mueller, Terry W. Ogletree
ISBN(s): 9780789728173, 0789728176
Edition: 4
File Details: PDF, 14.91 MB
Year: 2003
Language: english
Contents at a Glance
Part I:
Up Front: Network Planning and Design Concepts
Part II:
Physical Networking Components
Part III:
Low-Level Network Protocols
Part VI:
Fourth Edition LAN and WAN Network, Service, and Application
Protocols
Part VII:
Network User and Resource Management
Part VIII:
System and Network Security
Part IX:
Troubleshooting Networks
Part X:
Terry William Ogletree Upgrading Network Hardware
Part XI:
Migration and Integration
Appendixes
Index
On the Web:
Bridges, Repeaters, and Hubs
Token-Ring Networks
Fiber Distributed Data Interface (FDDI)
NetBIOS and NetBEUI
The Internet Message Access Protocol Version Installing the DHCP Server Service on
4 (IMAP4) 441 Windows 2000 or Server 2003 488
Transport Protocols 442 Authorizing the Server 489
Client Commands 442 Using the MMC Action Menu 490
System Flags 442 Configuring the DHCP Server and Scope
Retrieving the Message Header and Body Options 497
of the Message 443 Providing Support for BOOTP
Data Formatting 443 Clients 499
The User’s Inbox and Other Mailbox Enabling the DHCP Relay Agent 499
Naming 443 What Is a DHCP Cluster? 503
Universal Commands 443 Considerations for Using DHCP in Large
Other IMAP Commands 444 or Routed Environments 503
Non-Authenticated Commands 444 How DHCP Interacts with Microsoft’s
Authenticated Commands 444 Dynamic Domain Name Service
(DNS) 504
28 Troubleshooting Tools for TCP/IP Reservations and Exclusions 506
Networks 447 What Is APIPA? 507
Checking the Host System’s Configuration Troubleshooting Microsoft DHCP 508
First 448 Managing Logging 508
Using hostname and Related Using DHCP with Red Hat Linux 509
Commands 448 The DHCP Server Daemon 510
Using ipconfig and ifconfig to Check Host The DHCP Relay Agent 511
Configurations 449
30 Network Name Resolution 513
Using ping and tracert to Check
Connectivity 452 Hardware Versus Protocol Addresses 515
The ping Command 453 NetBIOS 515
The traceroute Command 457 The LMHOSTS File 515
The netstat and route Commands 461 Windows Internet Name Service 518
The arp Command 465 Installing and Configuring WINS on
The tcpdump Utility 466 Windows 2000/2003 Servers 524
The WinDump Utility 468 Managing the Windows 2000 WINS
Using the nslookup Command to Server 525
Troubleshoot Name Resolution Managing the Windows Server 2003 WINS
Issues 469 Service 529
Other Useful Commands 470 Using netsh Commands to Manage
WINS 530
29 BOOTP and Dynamic Host TCP/IP Names 531
Configuration Protocol The HOSTS File 533
Domain Name System 533
(DHCP) 473
Configuring DNS Clients 540
What Is BOOTP? 474 Using nslookup 540
Format of the BOOTP Packet 475 Dynamic DNS 541
The BOOTP Request/Reply Installing DNS on a Windows 2000 or 2003
Mechanism 476 Server 542
BOOTP Vendor-Specific Information Network Information Service 543
Options 477
Downloading an Operating System 480 31 Using the Active Directory 545
Taking BOOTP One Step Further: DHCP 480 Early Directories 546
The DHCP Packet Format and Additional The Difference Between the Directory and
Options 483 the Directory Service 547
The DHCP Client/Server Exchange 484 Interesting Objects 547
An Example: Installing and Configuring What Active Directory Delivers 548
a DHCP Server on Windows
2000/2003 488
x Contents This is the Chapter Title
From X.500 and DAP to the Lightweight Sequenced Packet Exchange (SPX) 590
Directory Access Protocol 549 SPX Packet Communications 591
The Active Directory Schema 551 SPX Packet Structure 591
Objects and Attributes 552 Sequenced Packet Exchange II
Standard Objects in the Active (SPXII) 592
Directory 553 NetWare Core Protocol (NCP) 593
What Is a Domain Tree? What Is a NCP Packet Signature Options 593
Forest? 554 Server Signature Levels 594
Domain Models—May They Rest in Client Signature Levels 595
Peace 555 Packet Signature and Job Servers 595
Partitioning the Active Directory into Effective Packet Signature Levels 596
Domains 555 Troubleshooting Packet Signature
A Domain Is Still a Domain 556 Conflicts 596
Active Directory Trees and Forests 556 NetWare Security Guidelines 597
The Active Directory and Dynamic DNS 557 NCP Protocol Independence 597
Dynamic DNS 558
How the Active Directory Uses DNS 558 33 Overview of the Novell Bindery
Using Sites to Manage Large and Novell Directory
Enterprises 559 Services 599
Directory Replication 560
Understanding NetWare Directory
Summarizing the Directory Data Using the
Structures 600
Global Catalog 561
Reviewing the Bindery Structure 600
Active Directory Service Interfaces
Reviewing the NetWare Directory Service
(ADSI) 561
Structure 600
Directory-Aware Application
Bindery Services 605
Programming 562
Contrasting and Comparing Bindery and
Now It’s Just Domain Controllers and
NDS 606
Member Servers 562
Using Novell Directory Services 610
The Active Directory Schema 563
Using NWADMN32 610
Modifying the Active Directory
Creating and Deleting Objects 611
Schema 563
Moving and Renaming Objects 613
Finding Objects in the Active Directory 571
Assigning Rights and Setting
Finding a User Account 571
Permissions 614
Finding a Printer in the Active
Using NDS Manager 619
Directory 574
Setting Up Bindery Services 622
Using Start/Search 575
Windows Server 2003: New Active Directory 34 Expanding and Enhancing NDS:
Features 575
NetWare’s eDirectory 623
Installing the Active Directory on a Windows
Server 2003 Computer 576 Basics of the eDirectory 624
The eDirectory Can Be Installed on Many
32 Overview of Novell NetWare Different Operating Systems 624
IPX/SPX 583 Options to Consider for Installing the
eDirectory 625
Using the Novell Proprietary Protocols 584
Hardware Requirements 625
The NetWare Protocol Suite 585
Installing the eDirectory for Supported
Connectionless Service and Protocols 586
Platforms 627
Connection-Oriented Service and
New Features the eDirectory Delivers 627
Protocols 587
TLS/SSL 627
Internetwork Packet Exchange (IPX) 587
iMonitor 628
IPX Packet Communications 588
The Simple Network Management
IPX Packet Structure 589
Protocol (SNMP) 628
IPX Frame Types 590
Extensible Match 629
Backup and Restore 629
This is the Current C–Head at the BOTTOM of the Page Contents xi
Index 1175
To my parents, Charles and Billie Jean Ogletree
And
Zira (1994–2002)
About the Authors
Terry William Ogletree is a consultant currently working in New Jersey. He has worked with
networked computer systems since 1980, starting out on Digital Equipment PDP computers and
OpenVMS-based VAX systems. He has worked with Unix and TCP/IP since 1985 and has been
involved with Windows NT and Windows 2000 since they first appeared, as well as the newest
additions to the family, Windows XP and the Windows Server 2003 family of servers. Besides
being the lead author of the third edition of this book, he is the author of Windows XP
Unleashed, Practical Firewalls, and The Complete Idiot’s Guide to Creating Your Own CDs (with co-
author Todd Brakke), and he has contributed chapters to many other books published by Que,
including Microsoft Windows 2000 Security Handbook and Special Edition Using Unix, Third Edition.
He is also the author of Fundamentals of Storage Area Networking. When not writing for Que, he
has on occasion contributed articles to PC Magazine.
You can email him at t@w2003tech.com or visit his home page at www.w2003tech.com. When
between jobs and not writing for Que, he can often be found on street corners holding a sign
that reads “Will work for hundreds of thousands of dollars.”
Dwight Tolay, Jr. (Chapter 6) started out as a computer test technician in the 1970s.
Branching out into the electrical construction industry, he became familiar with data and fiber-
optic cabling, has worked with coax Ethernet and IBM Token-Ring, and has followed the evolu-
tion up to today’s Category 6 and Gigabyte cabling methods. Currently, he is a general
supervisor for Ortlip Electric Co. He is a graduate EE, an ISA certified Level III control systems
technician, a licensed electrical contractor, and a certified high-voltage test technician. In addi-
tion to being a certified fiber-optic and teledata instructor at a local trade school for the past 13
years, he has contributed as technical editor on various books and currently is involved in a
book on Home Data and Electrical Systems Integration.
Scott and Kalinda Reeves (Chapters 32 and 33) are a married couple who live in Heron,
Montana, where they have written several networking exam books.
Scott has accrued his certifications as a Master Certified Novell Engineer (MCNE), Microsoft
Certified Professional (MCP) in Windows NT, Compaq Accredited Systems Engineer (ASE),
Comptia Network+ professional, and Comptia A+ certified technician. He has more than 15
years in the computer industry, and he has worked in the networking field for more than 11
years.
Kalinda has more than 16 years’ experience writing research, business, technical, and engineer-
ing documentation for government, military, and civilian customers. The topics include sys-
tem- and circuit-level hardware; uniquely developed, hardware-specific programs; and programs
that are implemented across government and military communications systems. Kalinda cur-
rently works as a freelance writer.
Acknowledgments
Most of the credit for getting this book done must be given not to the author, but to Rick
Kughen, Que’s executive editor, and to Todd Brakke and Mark Reddin, the development editors
of this book. There is no way I could ever have gotten this book finished without their consis-
tent, persistent, wonderful help. Todd Brakke has worked with me on four other books, and I
think both he and Mark Reddin deserve a lot of credit for the material you’ll find inside. Along
with Rick Kughen, they have contributed both questions and ideas about the new material that
we’ve included, as well as contributing to the organization of each chapter. Tricia Liebig
expended a lot of effort coordinating various people involved in this project—a job I would not
envy! As I am not the best writer in the world, I must also give credit to Cheri Clark, the copy
editor, for correcting my grammatical errors and for making helpful suggestions about better
ways to write this book’s text. Writing for Que is a team effort!
I also would like to acknowledge Sharon Terdeman, the Solutions Editor of PC Magazine. Several
of the articles she assisted me with spurred further research that resulted in more coverage of
those topics in this book. Additionally, Sharon is a good teacher when it comes to learning how
to write concise, informative text. She can take ten sentences and condense them into one, and
still make the same point. Helps a lot in a long book such as this one!
Most of my contracting jobs during the past five years have been the result of the efforts of
John Rogue and Angelo Simeo of The Computer Merchant firm in Norwell, Massachusetts. The
jobs they have been able to find for me not only pay the bills, but also have enabled me to
greatly further my knowledge of computers and networking. I can recommend this firm to any-
one who is looking for highly skilled employees or consultants in the computer and network-
ing fields (www.tcml.com).
A special thanks to Carl and Nanette Chiappetta for helping me stay focused on work and
enjoy life to its fullest. Thanks again to Jo and Jeff Johnson for being such good friends, and
the same for Jordan Scoggins, Andy Jones, James Garrett, Sari Gurney, Rick Clayton, and
Rodney Foster. Thanks also to Vicki Harding, my agent, for helping me get many writing
assignments (www.future-prod.com).
It goes without saying that without the help of Michael D. Parrott and Associates (and
Michael’s lovely and incredibly brilliant wife, Brenda), I never would have been able to find
time to write. MDP&A is the ultimate super-accounting firm that goes more than the extra mile
to take care of matters I just don’t have time for (www.mdp-a.com).
As an associate publisher for Que, I welcome your comments. You can fax, email, or write me
directly to let me know what you did or didn’t like about this book—as well as what we can do
to make our books stronger.
Please note that I cannot help you with technical problems related to the topic of this book, and that
due to the high volume of mail I receive, I might not be able to reply to every message.
When you write, please be sure to include this book’s title and author, as well as your name and
phone or fax number. I will carefully review your comments and share them with the author
and editors who worked on the book.
Email: feedback@quepublishing.com
For more information about this book or another Que title, visit our Web site at www.
quepublishing.com. Type the ISBN (excluding hyphens) or the title of a book in the Search
field to find the page you’re looking for.
Introduction
Since the last edition of this book, there have been many changes in the information and other
technology sectors. After the downturn in the dot-com industry, it looked as if a career in network-
ing might not be such a good choice after all. In the months leading up to the publication of this
book, however, things have started to change. The major baby bells and other large-scale network
providers have begun to expand their fiber networks and offer digital services to customers who
were not within their reach before. This is being fueled on the assumption that in the near future
voice services will be less of a revenue source than newer technologies such as end-to-end IP. XDSL
and cable modems have enabled both home users and business customers to connect even faster to
the Internet. These large corporations are laying the groundwork for what appears to be a very
bright future. If you think that networking and the Internet are futuristic now, just wait to see what
will happen tomorrow.
When put into perspective, the IT field grew dramatically leading up to the “year 2000” concerns.
And the continued growth of the Internet has been phenomenal—seemingly indifferent to the
economy at this time.
The growth in other IT fields, however, is historically unusual, and thus could not be sustained,
from a business standpoint. When investors are willing to throw money at just about any new
startup company because it seems like a safe ride, you can expect that there will be a backlash, as
has been the case. The downturn in the economy a few years ago forced many large companies to
put off purchases of network hardware, and also to lay off many employees. In general, the econ-
omy has always been a cyclic one, with a recession followed by a boom market. There is no reason
to expect this to change. But the good news is that the worst is probably behind us, and the future
is looking bright, especially because vendors have continued to develop new hardware and soft-
ware, just at a slower place. For example, look at how quickly wireless networking has grown,
despite the economy.
The next few years will most likely show a slower growth rate for IT jobs, but it will be a sustainable
growth; the future for networking jobs has never been brighter for the long term. Because of my
belief in this trend, it was much easier to write this fourth edition of Scott Mueller’s Upgrading and
Repairing Networks. I was enthusiastic because over time I have been able to add new topics,
expand on others, and move some older material (which may still be relevant to your job) to the
upgradingandrepairingpcs.com Web site that accompanies this book. I believe that with each new
edition of this book, it just gets better and better. And with great development and technical editors
at Que helping to make this a better book, I think you will find it the most comprehensive book on
networking on the market today.
Part III, “Low-Level Network Protocols,” is where you’ll find information about many types of tech-
nologies you can use to transmit data across your network. This section starts with a quick review
of the IEEE 802.11 standards, followed by a chapter on the oldest LAN technology still in wide-
spread use today: ARCnet. If you think that older solutions are always supplanted by newer ones
that perform the same functions more efficiently, then read that chapter. And the next time some-
one asks you whether you want fries with that burger, you just might find that ARCnet is still
there, as a viable solution to small networks that cannot afford a full-time network administrator.
Because of ARCnet’s simplicity, it is still employed in various situations, such as in point-of-sale reg-
isters and on some factory floors.
Today the major LAN wire protocol is Ethernet. It’s been here for years, and will continue to be
around as it evolves to meet newer demands for bandwidth and adapts to newer devices and tech-
nologies.
Token-Ring is still being used by a small percentage of networks, supported mainly by IBM and a
few other vendors. However, because of its miniscule market share, and because there is very little
development in process to update the technology as compared to Ethernet, this updated chapter is
now found on the upgradingandrepairingpcs.com Web site.
Part IV, “Dedicated Connections and WAN Protocols,” is written for both network administrators
who need to connect LANs across large distances and those who need a local dedicated connection
to another LAN or the Internet. Here you can find the choices available to you from the high-end
T-class connections that give you huge bandwidth data paths. You will also find information about
other dedicated connections, such as cable and DSL access. Keep in mind that cable and DSL access
are not just for home users. Many of the large providers of these services are now targeting small
businesses. If this is your environment, you might find this a less expensive method for giving your
company an Internet presence.
Part V, “Wireless Networking Protocols,” covers just what it says. In the previous edition of this
book, wireless networking was just making its way into the marketplace. Even Bluetooth—used for
very close connections, such as to replace cables used between keyboards, mice, and other devices
to your computer—has probably finally arrived. In the past few years there has been much market-
ing hype about this technology, and it finally seems to have made it. You can also learn here about
the major wireless technologies—many available at the consumer level as well as for the corporate
network—including both IEEE 802.11b and IEEE 802.11a. The newest protocol in this section is
IEEE 802.11g, which combines the best of both of the other two protocols. The 802.11g network
devices are backward compatible with IEEE 802.11a and b, and can offer bandwidths up to 54Mbps.
For wireless networks, you won’t notice much of a difference from this sort of connection to your
company’s network than if you had a 100Mbps 100BASE-T connection. To top off this section,
there is a chapter on other wireless developments, including everything from wireless security to
new uses for wireless technology.
Part VI, “LAN and WAN Network, Service, and Application Protocols,” contains such subjects as the
all-important TCP/IP suite of protocols, applications, and troubleshooting utilities. Because TCP/IP
is the main protocol used in LANs today, this is a must-read section for new readers as well as a
refresher course for those who already are familiar with the topic. In addition to the basics of
TCP/IP, this section covers everything from email protocols to directory services, routing protocols,
and SSL, among many others.
4 Introduction
Part VII, “Network User and Resource Management,” will help you learn about managing users and
controlling access to network resources. Coverage here includes Windows, NetWare, and
Unix/Linux. One of the newest additions to this section is Windows 2003 and how you can use it
and the updated Active Directory to manage and secure the network. A chapter devoted to network
printing protocols has been updated to include the latest coverage of the Internet Printing Protocol
(IPP), which is now supported by all the major operating systems. This is a topic that will revolu-
tionize printing in the next few years, and it’s a good idea to start brushing up on your skills today.
Part VIII, “System and Network Security,” contains chapters to help you learn the basics about pro-
tecting your system. Topics include everything from creating policies and procedures for the work-
place, to computer viruses and auditing the actions users perform on the network. Encryption
technology is also covered in its own chapter, as are Virtual Private Networks (VPNs) and firewalls.
If you connect to the Internet, or if you operate a large enterprise network, then this section has a
lot of good security issues you should be cognizant of.
Part IX, “Troubleshooting Networks,” complements most of the previous chapters. Tools you can
use for troubleshooting and strategies for going about this process are found here. For small
office/home office (SOHO) network users, there is an entire chapter devoted to your network.
Because many SOHO networks are operated by small business owners, this chapter can help you
understand more complex topics without having to read this entire book. Off-the-shelf cable/DSL
routers and inexpensive firewall solutions might make you feel comfortable about your network,
but everyday there are new threats that can make your small network a target.
Part X, “Upgrading Network Hardware,” is a valuable reference for those who want to make use of
the latest hardware devices, while trying to protect your investment in existing equipment where
possible. A special chapter gives a good overview of adding wireless networking to your wired net-
work.
Part XI, “Migration and Integration,” covers the other side of the coin, upgrading from one operat-
ing system to another, as well as using multiple operating systems on the same network. This sec-
tion discusses the many things that are common to most operating systems, as well as tools that
can be used to help you integrate a diverse collection of systems into a single heterogeneous net-
work. Topping off this section is another chapter of interest to SOHO users that discusses how you
might go about setting up a LAN for your business. Again, you don’t need to be versed in all the
topics covered in this book in order to operate a small LAN. This chapter does refer to topics in
other parts of the book, but it is a good starting place for those who either have a small LAN that
needs to be upgraded, or are just starting out.
Finally, the appendixes contain information that can be used as a quick reference. You can get an
overview of the OSI network reference model, as well as a concise explanation of how directory ser-
vices work, specifically the Lightweight Directory Access Protocol (LDAP). An important resource
here is a comprehensive glossary of networking terms. If you are reading a chapter in this book and
find yourself stymied about a term or an acronym, then just look to this appendix for help.
As discussed earlier in this introduction, the upgradingandrepairingpcs.com Web site contains sev-
eral chapters that had to be moved out of the main text. This is due to several factors. First, some of
these topics apply to older networks and, second, this book would cost you a lot more if the page
count were to increase by a few hundred pages! The chapters on the Web site, however, should not
be dismissed as irrelevant. You may have a network that still uses the technologies found in these
chapters. And, when studying for a certification exam, you should understand older technologies
because they will probably turn up as questions when you sit down to take the test.
Introduction 5
PART I
1
A Short History of
Computer Networking
CHAPTER 1
10 Chapter 1 A Short History of Computer Networking
Today, computer networks are taken for granted much as the telephone network is. And the tele-
phone network was, until the explosive growth of the Internet, the largest network in the world. It
just wasn’t a computer network. You could use modems to connect computers to each other on a one-
by-one basis, but this wasn’t networking in the sense we think of it today. And those early modems—
300 baud or less—didn’t make the transfer of data an inexpensive matter, especially when
long-distance calls were required. In a funny twist of fate, voice communications are now creeping
into the networking world (such as voice over IP, or VOIP), making the telephone network itself less
important. To that end, many telephone companies are expanding rapidly into the data networking
field so that they can offer data, voice, video, and other services.
Computer networking has been evolving since the late 1960s. Early work on the ARPANET began in
the 1960s, and in 1969 a four-node network using primitive packet switching was created. The growth
of this predecessor of today’s Internet wasn’t quite the phenomenon you see today.
As discussed in Chapter 14, “Ethernet: The Universal Standard,” the ALOHAnet was created to estab-
lish connections between several computers in Hawaii. Robert Metcalf was later to use the basic prin-
ciples from ALOHAnet to create what eventually became the Ethernet local area networking wire
protocol still used today. Ethernet was simply a means to get a signal from one place to another—it
was another thing to decide what kind of signaling to use. On the ARPANET, TCP was being created,
and refined. If you examine the OSI Seven-Layer Networking Reference Model (see Appendix A,
“Overview of the OSI Seven-Layer Networking Reference Model”), you can see that Ethernet works at
a low level in the model, and is used to transmit packets of information from higher-level protocols.
TCP was the first major higher-level protocol created. One of the first refinements of TCP was to break
it into several parts (or layers), which is why the protocol “suite” is known today as TCP/IP.
TCP (the Transmission Control Protocol) and IP (the Internet Protocol) are the basis of the Internet.
Another break-out from TCP was the User Datagram Protocol (UDP). Both TCP and UDP provide dif-
ferent types of service, yet both use IP as the workhorse protocol that is used to route packets (or data-
grams) on the Internet, as well as any intranet. And what does IP use to send data across the wire (or
the air, in case of a wireless network)? Ethernet in most cases. For the long-haul, there are other proto-
cols that can be used, and you will read about them in Part V of this book, “Wireless Networking
Protocols.”
It is important to keep in mind that Ethernet and other wire-level protocols simply provide the means
to frame data—create discrete units of data for transmission—and then use a specified method to send
data across the network media. For example, simply varying the voltage on a wire can be used to send
data from one point to another. An early transmission method, called non-return to zero (NRZ)
encoding, used just this method. A high value was used to specify the bit value of one, and a low
value was used for zero. A drawback to this encoding scheme is that a long stream of either ones or
zeros can be difficult to decipher at the receiving end. The term “clock” is used to mean that each end
of the transmission understands where a bit starts and where it ends during the transmission. Because
it would be very expensive to have a physical clock at each end that could precisely time each bit
transmission, this encoding method does not scale well.
Early Ethernet networks used a technique called Manchester encoding. This method does provide a
clocking mechanism that is built into the coding scheme itself. Instead of using a high- or low-voltage
state to indicate a specific bit, Manchester encoding uses the change from one state to another, during
a specific interval.
While TCP/IP continued its development on the ARPANET, computer vendors began to recognize the
importance of networking, and many proprietary protocols were developed. Digital Equipment
Corporation (DEC) created DECnet (and numerous other protocols), which was used to connect its
PDP computers and, later, VAX and AlphaServer computers. Today DECnet is still used, although
A Short History of Computer Networking Chapter 1 11
TCP/IP has pretty much replaced it for most installations. For a short time in the 1980s, however,
DEC operated the largest computer network in the world, short of the Internet. What protocol was
used? DECnet, of course. During that same period, the OSI model was created, and Digital incorpo-
rated the concepts of that model, as well as the protocols that were developed by ISO based on the
OSI model, into DECnet. Because the VMS (Virtual Memory System) operating system used on DEC’s
VAX computers was adopting these open standards, the name of the operating system was changed to
OpenVMS. However, few other vendors chose to incorporate the high-overhead concepts of these
open protocols, and this first attempt to standardize networking protocols between different comput-
ers failed to come about.
Other computer manufacturers also produced their own proprietary network protocols. For example,
IBM’s work in this direction resulted in SNA, which combined networking protocols from the high
end (application) down to the low end (wire protocols). As networking began to become an important
part of the computing world, other vendors, such as Xerox (XNS), also came up with their own proto-
cols.
The result was that if you wanted to create a network of computers for your business, you had to stick
with a single vendor. Proprietary protocols, then, were not a good solution to the problem of
exchanging data between computers.
During the early days of PCs, the same sort of situation occurred. Although PCs were basically the
same when it came to the operating system (DOS at the start), you could buy a network setup from
many different vendors. One that comes to mind is NetWare (which is still around today, although in
recent years TCP/IP has replaced the proprietary IPX/SPX NetWare protocols). LAN Manager was
Microsoft’s entry into the field, with a legacy of NetBEUI and NetBIOS still lurking around on many
Windows computers prior to Windows 2000. When DEC started to build its own PCs, it licensed LAN
Manager technology and created Pathworks. You might still find Pathworks in some networks,
although, like LAN Manager, it is considered history today.
Other networking packages included Banyan Vines and LANtastic, both of which are still around
today. However, these products are today vastly different than when they first appeared.
In Chapter 13, “The Oldest LAN Protocol Is Still Kicking: ARCnet,” you can read about an old proto-
col that is still used pretty much the same as it was when first created. It allows a limited number of
computers to be connected, uses a simple token passing scheme, and requires minimal setup. ARCnet
is typically found in point-of-sale computers and factory automation today.
The Internet changed the entire landscape. As TCP/IP continued to mature into the stable protocol
suite that it is today, the PC landscape, as well as mainframe and minicomputers, began to adopt
TCP/IP. Although the ISO first attempted to define open protocols so that computers from different
vendors could interact to exchange data, it turns out that TCP/IP is the winner in the end. And when
IPv6 (IP version 6) finally reaches from the inner core of the Internet to the edge, you will find that
TCP/IP continues to add new features, enhance security, and provide more robust features.
Other protocols, such as ATM and Frame Relay, are used for long-distance transfer of data, and can
encapsulate other protocols such as TCP/IP. Fibre Channel is a wire protocol that is the most widely
used protocol in Storage Area Networks today (see Chapter 11, “Network Attached Storage and Storage
Area Networks”).
The old standard Ethernet has itself continued to be enhanced to keep up with the need for speed.
Early versions operated at 2–5Mbps, and most desktops today use 100BASE-T, or 100Mbps Ethernet.
Gigabit and 10Gigabit Ethernet are now on the market, although these newer versions do not use the
same signaling techniques as earlier versions. The capability to provide backward compatibility with
earlier versions, however, is another important factor for the continued use of Ethernet.
12 Chapter 1 A Short History of Computer Networking
Today you will find that most desktop computers in a company’s LAN use TCP/IP. Although other
protocols may encapsulate TCP/IP for transmission over a long distance, the TCP/IP protocol is still
the de facto standard for computer-to-computer communications. TCP/IP is also supported by net-
worked printers and wireless communications.
Because of this standardization, prices for equipment that support Ethernet and TCP/IP are dramati-
cally less than a decade ago. Network adapters themselves may become history because many com-
puter motherboard manufacturers are starting to incorporate that functionality directly onto the
motherboard.
So what does this all mean? It means that whether you operate a business or a home network, or if
you just connect to the Internet from home, it has been a long process to get to where we are today.
In this book you will find topics that cover many of the important protocols in use today, as well as
topics on newer developments.
2
Overview of Network
Topologies
LAN Topologies 14
Building and Campus Topologies 24
A Multi-Tiered Network Topology 28
CHAPTER 2
14 Chapter 2 Overview of Network Topologies
Before you can begin to upgrade and repair your network, you need to understand how it’s laid out,
how it functions, and how the various parts are related to one another. Knowing how your network
components are related makes the extension, expansion, and troubleshooting of your environment
more focused and productive. Because network uptime is related directly to productivity, a solid grasp
of network concepts is a necessity when you’re facing a troubled LAN.
In this chapter, you will review the topologies in use today and learn the strengths and weaknesses of
each.
LAN Topologies
Several unique network technologies have been developed over the past three decades. Different types
of networks have different design criteria and, thus, various topologies have come into use. One
important distinction needs to be made before we enter into a serious discussion on topology: physi-
cal topology versus logical topology. The physical topology describes the layout of a network media
(such as copper and fiber-optic cables and, more recently, wireless equipment) and the devices that
connect to it. The logical topology is concerned not with the actual physical connections but with the
logical path through the network that data can take from one place to another. The differences will be
more evident as the different topologies are discussed.
The basic topologies you will find in most LANs today include the following:
■ Bus
■ Star
■ Ring
■ Mesh
■ Hybrids
Bus Topology
The simple bus topology structure was the first type used in Ethernet networks. The typical bus physi-
cal topology consists of a coaxial cabling common to all computer systems connected to the LAN.
This coax is tapped in multiple places along its length, with each tap being used as a point of connec-
tion for a computing system. Taps can be physical cores cut into the coax (sometimes called a “vam-
pire tap”) or BNC-style “T-connectors” that join several individual pieces of coax together to form the
common bus (see Figure 2.1 and Figure 2.2 for a comparison of the two methods).
Vampire Taps
Terminator
Terminator
Figure 2.1 Computers can connect to coaxial cables on a bus by tapping directly through the core of the
cable. The vampire taps pierce the thicknet cable but not the BNC.
LAN Topologies Chapter 2 15
Figure 2.2 BNC-style T-connectors make attaching computers to a bus a simpler operation.
◊◊ You can learn more about how 10BASE-2 and 10BASE-5 Ethernet networks are created using coaxial cables
and the bus topology in Chapter 14, “Ethernet: The Universal Standard.”
A bus is also a logical topology. From a device’s viewpoint, all other systems communicate through
the same, shared path. Because it is a shared media technology, mechanisms must be put into place to
arbitrate network traffic over the cable. Typically, collision detection (CD) or collision avoidance (CA)
algorithms are used in bus topologies to arbitrate network access along with concepts such as “broad-
casts” to reach every device on the cable. This subject is covered in detail in Chapter 14.
The bus topology is very simple and inexpensive to implement due to its low cost requirements for
cable installation (there’s only one main trunk). But some serious deficiencies make bus topology
LANs unattractive to deploy:
■ Bus topologies require proper terminations on both ends of the bus to effectively dampen the
network signal and to avoid a “reflection” or reoccurrence of a previous transmission. Without
the proper terminations in place, expect a very slow or inoperable network.
■ The cable itself is a single point of failure. One break, cut, or poor connection negatively
impacts the entire LAN.
■ Because all workstations or devices share a common cable, troubleshooting can be difficult
when problems occur. You must temporarily break the terminations in the network to isolate a
device. After you think you’ve resolved the network problem, you must disrupt LAN service
again to reattach the device to the network. This makes for a cumbersome and disruptive
process.
Due to these limitations, the bus topology is typically found only in the smallest or most austere of
installations. Some proprietary manufacturing process control systems use a bus topology, but these
aren’t covered in this book.
For the most part, the bus topology is a historical relic. However, it is something you should be cog-
nizant of in order to understand why other topologies are the norm today. Early networks were com-
posed of only a few computers and there was no need to provide for today’s high-bandwidth
networks that use switches and other devices to connect a diverse collection of computers and other
networked devices.
Star Topology
The concept behind the star topology is simple. Every node on the LAN has a dedicated cable that is
pulled back to a centralized point, typically a wiring closet. All cables are terminated in a network
component within the closet, such as a hub or, more typically today, a switch, which handles the
repeating or switching of traffic out to the other nodes on the network (see Figure 2.3).
16 Chapter 2 Overview of Network Topologies
Hub/Switch
Figure 2.3 The star topology enables you to centralize wiring for a network.
The shortcomings of a star topology network are obvious: The network component (hub or switch) is
a single point of failure, and a great deal of wiring is involved to implement the star.
However, there are tremendous benefits to a star topology:
■ Management of the network is centralized around the hub and switch components. Most of
these components have features that allow an administrator to spot congestion and network
errors at the port level, which makes troubleshooting problems quick and simple.
■ “Smart hubs” can automatically disable ports that exceed use or error thresholds, providing
additional stability to the LAN. They are also a central point for watching bandwidth usage and
overall network health.
■ Wiring installation is less obtrusive and therefore does not disrupt LAN service with the addi-
tion or deletion of nodes.
■ A cable cut or bad connector does not take down the entire LAN segment. No terminators are
required as in the bus topology model.
You should note that nearly every popular network technology today uses a star topology for its phys-
ical implementation. This is due to several factors, including the ease of wiring, the fact that a single
misbehaving computer can be removed from the topology, and the fact that it’s a simple matter to set
up a hub or switch.
Note
When troubleshooting a star topology network, be sure to check error counters and status indicators on your network com-
ponents. These can provide valuable information in helping you find what is at fault.
Ring Topology
Ring topologies are more complex than the bus and star topologies discussed in previous sections, but
they offer some attractive features. Nodes logically communicate in a ring formation, with each node
communicating only directly with its upstream and downstream neighbors (see Figure 2.4).
LAN Topologies Chapter 2 17
D B
Figure 2.4 The ring topology links each node on the network to two other nodes on the network.
You can probably picture what a mess the wiring would be in an office with hundreds of computer
systems, if a network like this were implemented as a physical ring. So ring topologies are typically
implemented in a physical star topology (see Figure 2.5).
In a ring topology, access to the network is controlled through a token that is passed from node to
node as the arbitration mechanism. Each node takes its turn at claiming the token as the token passes
from neighbor to neighbor, and when a node possesses the token, it takes its turn to transmit onto
the ring. A data packet is transferred from one node to the next until it reaches its destination node.
After the destination node has received the packet, it modifies the packet to acknowledge receipt and
passes it on. Eventually, the packet makes it completely around the ring, and the transmitting node
receives it and notes that the receipt has been acknowledged. When the transmitting node is finished,
it releases the token to its neighbor, and the process repeats.
Note
Token-Ring networks are the primary LAN technology that uses a ring topology. Although Token-Ring technology today rep-
resents only a small percentage of network installations, another network, the Storage Area Network (SAN) still uses the
concept of a ring topology, with a different method for gaining access to the network media than the method used by
Token-Ring networks.
In Chapter 11, “Network Attached Storage and Storage Area Networks,” you will find that Arbitrated Loops continue to
use the ring topology to provide access to storage devices, such as disk and tape drives, to high-end servers. The
18 Chapter 2 Overview of Network Topologies
Arbitrated Loop maintains a ring topology but uses an arbitration priority method based on device addresses to gain
access to the loop instead of the token-passing mechanism used by Token-Ring networks.
Newer SANs are built using a switched network technology often referred to as a SAN fabric. As you will read in
Chapter 11, it is possible to attach an Arbitrated Loop to a SAN fabric to preserve your current investment.
Flow of Communication
Physical
Wiring
Figure 2.5 The ring topology is usually implemented as a physical star to simplify wiring management.
Another ring topology called Fiber Distributed Data Interface (FDDI) uses dual fault-tolerant rings.
This technology requires that the two rings have tokens passing in opposite directions of one another.
A breakdown in one ring causes the nodes to shift over to the secondary ring to continue communi-
cations.
◊◊ FDDI is another networking technology that uses the ring topology and token-passing for media access. Please
visit the upgradingandrepairingpcs.com Web site for the chapter “Fiber Distributed Data Interface (FDDI),”
which covers this topic in detail.
The down side of a ring topology is simple: The firmware required to manage the ring is somewhat
complicated and must be on every network card that participates in the ring. Because Ethernet domi-
nates the marketplace today, you can purchase an inexpensive 100Mbps Ethernet network adapter for
$20 to $30. Token-Ring network adapters are a bit more difficult to find, and they tend to be more
expensive devices than their Ethernet counterparts. As a result, technologies such as Ethernet have
transitioned rapidly to higher networking speeds while ring topologies have never quite jumped the
speed gap. The ring topologies that exist today have changed little in the past 10 to 15 years.
LAN Topologies Chapter 2 19
Note
In Chapter 12, “The IEEE LAN/MAN Committee Networking Standards,” you will see that development on Token-Ring
networks has pretty much come to an end-of-life state as far as standards are concerned. Although Token-Ring networks
are still used in some vertical market applications (such as the factory floor), Ethernet technology has become the norm for
the LAN. If you want to learn more about Token-Ring networks, you can visit the primary vendor of this technology at
www.networking.ibm.com/.
Mesh Topology
A mesh topology is an interlacing of multiple connections among several nodes. Typically, a mesh is
done for one purpose: redundancy. Any serious campus network must incorporate a mesh to achieve
the level of redundancy and fault tolerance that businesses demand from their data networks. There
are two types of mesh: full and partial.
Except in the smallest network, a full mesh is not very practical, but it is mentioned here for com-
pleteness. Full mesh means that every node contained in a network has a connection to every other
node contained in the network. It should be fairly obvious at this point why full meshes are not very
practical (see Figure 2.6). The cost for such infrastructure would be exorbitant, and 90% of it would
never be put to use.
Figure 2.6 A full mesh topology is not a very practical way to wire a network.
Partial meshes are designed to provide redundancy where it is needed. By using a little forethought in
design, a network architect could place some additional connectivity where it can provide needed
bandwidth and fault tolerance to the network. Suppose for a moment that an important resource
were attached to node A, as shown in Figure 2.7.
You can spot several paths that could fail, yet all the nodes of your network would still be able to
reach node A as a destination. The true merits of partial mesh are realized when you look at WAN and
campus topologies.
Hybrid Topologies
Hubs or switches can be attached to one another to create larger LANs capable of supporting more
devices. After this happens, you start getting some interesting hybrid topologies. Three popular
hybrids are tree, hierarchical star, and star-wireless.
20 Chapter 2 Overview of Network Topologies
Figure 2.7 A partial mesh topology can be used to provide redundancy for the network.
Tree
Figure 2.8 shows a combination topology that groups workstations in a star and joins the stars along a
linear bus. The majority of the problems of the bus are eliminated because a single workstation can-
not bring the entire LAN to a halt. You still can add or change workstations by plugging them into a
different port on the same hub, or on another hub. If one hub malfunctions, it disables only the
workstations that are attached to it from communicating on the network. The remaining workstations
on the other hubs can continue to function normally.
This is an inexpensive method that can be used to join different work departments in a building.
Each local workgroup can have an administrative person who is responsible for managing the connec-
tions on the local hub. The network administrator can regulate when and where new hubs are
attached to the network. This also can be used to help extend the distance of a LAN. For example, you
can use 10BASE-2 cabling to connect two 10BASE-T networks that are in separate buildings. However,
today that connection would most likely be accomplished using more modern techniques, such as
with fiber-optic cabling.
The major problem with this type of hybrid topology, however, is that if there is a problem with the
backbone bus cable, the network becomes segmented into individual hubs. Workstations on each hub
can communicate with each other, but data transfers through the network to workstations on other
hubs will be disrupted until the cable problem is diagnosed and corrected.
Hierarchical Star
Another method that can be used to connect hubs is a hierarchical star. This method, shown in Figure
2.9, uses a central hub to link several hubs that have workstations attached.
This method can be used to build very large LANs; however, there are restrictions on the size of the
LAN. Timing issues as well as address space are driving factors in how many hubs or switches you are
able to attach in the hierarchical star topology without the introduction of routing technology. The
various restrictions of different network technologies are discussed in later chapters.
LAN Topologies Chapter 2 21
Node
Node Node
Hub Hub
Node Node Node Node
Node Node
Figure 2.8 A combination of the bus and star topologies groups workstations in a star and joins them
along a linear bus.
Hub
Hub Hub
Node Node
Star-Wireless
The most recent hybrid topology has arrived with the advent of wireless technology. Wireless LAN
technology in its current implementation requires a user to be in the vicinity of an access point
attached to the wired data network. A configuration such as this gives you an amorphous hybrid
topology of star combined with wireless (see Figure 2.10).
The star topology is necessary to combine the many access points spread across a building to ensure
wireless coverage. All the access points collapse back into the main star hub, where server resources
would reside on the network. As wireless technologies continue to evolve, so will the topologies that
support them.
22 Chapter 2 Overview of Network Topologies
Switch/Hub
Switch/Hub Switch/Hub
Switch/Hub
Wireless Wireless
Wireless
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookultra.com