Home Networking Bible 2nd phần 9 pps

In order to use the System Monitor over the network, you need the following: ✦ A client/server network using either NT Server or Novell NetWare server ✦ User-level security selected for

Another simple thing to check is to see if the user has changed any network configurations,added any hardware or software, or had any problems with the computer lately If you askthose questions first, you might get to the root of the problem more quickly.

Using the System Monitor

If you have network problems and you use a client/server network, you can use the SystemMonitor to help find the cause of the problems on a Windows 98 computer If you don’t use aclient/server network, you can view one machine’s use of system resources to help find thebottleneck in your system The System Monitor enables you to monitor running processes,memory usage, dial-up access, and more

In order to use the System Monitor over the network, you need the following:

✦ A client/server network using either NT Server or Novell NetWare server

✦ User-level security selected for each computer, which removes all network shares youset for a peer-to-peer workgroup

✦ The Remote Registry Service installed on each computer

✦ File and Printer Sharing enabled on each computer

Using the Network Troubleshooter

Windows XP has included a network troubleshooter that can help you determine problemswith adapters, terminal services, modem sharing, even TCP/IP configuration To use theNetwork Troubleshooter, right-click My Network Places and choose Properties In theNetwork Connections dialog box, click Network Troubleshooter

The Network Troubleshooter performs many tasks It can perform tests on TCP/IP by forming the PING command and/or the NETVIEW command for you The Network Trouble-shooter can check your dial-up or Internet connection as well, automatically

per-If you use Windows 2000, you can use Windows Help to locate a variety of networkingtroubleshooters One, in particular, is for TCP/IP networks; another is for users; another forhardware; and so on

Examining the Network Hardware and Software

Network connection problems sometimes affect only one computer, but they often affect allcomputers on the network When you have a problem, you can check a computer’s networkinterface card, cabling, and then the network hub

You also can check the protocols used on the network TCP/IP configuration often causesconnection problems as well A problem with network speed might have to do with hardwareand software If you have problems with bottlenecks, you should check certain areas of thenetwork

When experiencing network connection problems, ask yourself the following questions tohelp diagnose the problem:

Note

✦ Did the connection work before, or did it just recently stop working?

✦ If it just recently stopped working, what equipment or software have you added?

✦ Did something happen or change since it last worked?

If the connection never worked, you probably have configuration problems or a bad cable If

it recently stopped working and you added something new to the network, remove that

added equipment and see if the situation improves If nothing’s changed and the connection

quit working all of a sudden, check network cards, cabling, and hubs, in that order

If you have a problem you cannot solve on your own, check the Microsoft Knowledge Base

for help The address is www.support.microsoft.com

Exploring network card problems

If your computer won’t connect to the network, you might have a problem with your network

card If the computer is new and you just added the card, check the IRQ (interrupt request)

and DMA (direct memory access) addresses Configuration might be the problem If the card

worked and then just stopped working, the card might need to be replaced

The easiest and quickest troubleshooting solution for a network that is not connecting is to

delete everything from the network properties dialog box — TCP/IP, client, services, and

net-work adapter — reboot, and then install everything anew This often refreshes corrupted

soft-ware or settings and makes the network magically work

Checking in the Device Manager

If you’re having a connection problem, check to make sure that Windows sees your network

interface card If you see a red or yellow icon to the left of the NIC, then there’s a problem

with your card Figure A-6 illustrates the Device Manager tab of the System Properties dialog

box To check the adapter card in the Device Manager, follow these steps:

Figure A-6: Check your system devices.

Tip

Tip

1 Choose Start ➪ Settings ➪ Control Panel ➪ System in Windows 98 In Windows 2000 and

XP, right-click My Computer ➪ Properties; click the Hardware tab and then DeviceManager The System Properties dialog box appears

2 Choose the Device Manager tab in Windows 98, and choose the View devices by type

option

3 Click the plus (+) sign to the left of Network adapters.

4 Check to see if your adapter has a yellow exclamation point in a circle or a red x The

yellow exclamation point means the device isn’t working properly; the red x means thedevice isn’t working at all

5 If your adapter card does display one of these icons, select the adapter and then click

the Properties button The adapter card’s Properties dialog box appears, as shown inFigure A-7

Figure A-7: Look at the adapter’s

properties

6 In the General tab, view the Device status area for a definition of the problem.

7 Click OK to close the Properties dialog box.

8 Close the System Properties dialog box, and then close the Control Panel.

Some network cards, especially the Plug and Play cards, come with special software that youcan use to configure the card If you’re having trouble with the IRQ or other addressing infor-mation, check to see if your card manufacturer supplied this software

Checking network settings

You also should verify that all your settings in the Network dialog box are correct To checkyour adapter settings, follow these steps:

1 Choose Start ➪ Settings ➪ Control Panel in Windows 98 The Control Panel window

appears In Windows 2000 or XP, right-click My Network Places ➪ Properties

Note

2 Double-click the Network icon in Windows 98 In Windows 2000 or XP, right-click Local

Area Connection and Properties The Network (Windows 98) or the Local Area

Connection Properties (Windows 2000 and XP) dialog box appears

3 In the Configuration tab, select your network adapter card, as shown in Figure A-8 in

Windows 98 In Windows 2000 or XP, click the Configure button

Figure A-8: Select your network card.

4 Click the Properties button in Windows 98 or the Resources tab in Windows 2000 or XP.

The adapter’s Properties dialog box appears

5 Choose the Resources tab if you’re using Windows 98, as shown in Figure A-9.

Figure A-9: Check the IRQ and I/O

address range

For more information about IRQs and I/O addresses, see Chapter 25.

6 Make sure the configuration is correct, and then click OK.

7 In the Configuration tab (Windows 98) or the General tab (Windows 2000 or XP), check

to be sure you have the Client for Microsoft Networks installed

8 Next, check the Identification tab of the Network dialog box in Windows 98, as shown in

Figure A-10, to make sure that you’re using the correct workgroup name and that yourcomputer name is unique in the network For Windows XP, you must right-click MyComputer ➪ Properties and choose the Computer Name tab For Windows 2000, right-click My Computer ➪ Properties and choose the Network Identification tab The com-puter name and workgroup name are listed You can change either by clicking theProperties button

Figure A-10: Make sure you have the

right workgroup name

9 Click OK Restart the computer if prompted.

If your power management is on, try turning it off Sometimes power management featurescan cause network problems In the Control Panel, double-click the Power Managementicon In Settings for Always On (Power Scheme area), choose Never for the following twoitems: Turn Off Monitor and Turn Off Hard Disks

Looking at protocol problems

When two or more PCs have problems communicating with each other, you could have a tocol mismatch Say that one PC cannot see others on the network Check that PC’s protocolsettings and bindings Compare the protocols listed in the Network dialog box of a workingcomputer to the one that isn’t connecting Check all protocol properties as well

pro-Tip

Cross-Reference

For information about protocols, see Chapter 10 or Appendix B.

If you’re having trouble with network connections, it could be the protocol When running

TCP/IP, configuration is usually the first thing you should check Use PING, an MS-DOS prompt

command you use to test TCP/IP configuration, to verify that the TCP/IP configuration is

cor-rect, that local computers are communicating with each other, and that remote computers

are communicating For more information, see Appendix B

Use IPCONFIG to display the IP addresses, subnet masks, and default gateways for all

net-work adapter cards on each computer Make sure that the IP addresses are unique, that

sub-net masks match, and so on IPCONFIG is a command you use to test IP addressing For more

information, see Appendix B

Use NETSTAT to track down strange or unusual network problems You can troubleshoot

incoming and outgoing packet errors with this command NETSTAT is an MS-DOS prompt

com-mand you use to test TCP/IP configuration For more information, see Appendix B

Use WINIPCFG, an MS-DOS prompt command you use to test TCP/IP configuration, if you’re

having trouble with your Dynamic Host Configuration Protocol (DHCP) lease DHCP is a utility

for assigning TCP/IP addresses to workstations automatically Many DHCP servers provide

networked computers with a lease that enables them to use the same number for a limited

amount of time You can use this command to release a lease when you stop using DHCP or

renew a lease that you want to continue For more information, see Appendix B

See Appendix B for more information about troubleshooting techniques to use with a TCP/IP

network

Troubleshooting cabling problems

You might be using coax, twisted-pair, phone line, or power line cables for your network For

reasons of safety, problems with your power line are best taken care of by an electrician Phone

line problems may require a professional as well, although you can check some of the phone

line yourself

If you didn’t install your own coax or twisted-pair cabling, you might want to call for help in

checking the cabling But you also can check some problems yourself

For more information about coax and twisted-pair cables, see Chapter 6 For more

informa-tion about phone and power line network cabling, see Chapter 8

Looking at the phone line

You can check the phone wiring that connects the computer to the wall jack for nicks, crimps,

and other problems If you see an obvious problem, replace the wiring You can buy a short

length of phone wire, with the RJ-11 connectors attached, at a hardware store, discount

depart-ment store, or any place that sells telephones

If you don’t see a problem but still think it might be the phone wiring, try trading the current

wiring with another phone line wire to see if it works You then can replace the bad wire if

that will solve the problem

Considering coaxial cable

Check all of the cable in the system for nicks or breaks first Also check all the terminators tomake sure that they are securely fastened to the cable and to the computers Even if all looksokay, you still might have a break in the connection; it’s difficult to tell with coaxial cabling.You can try replacing the terminators and T-connectors, if you’re comfortable with that Youalso can try swapping out one cable with another too, unless your cabling goes into walls orunder flooring

You also can use a cable tester, if you have one Cable testers are devices that test for loose

connections, faulty cables, and other cabling problems You can buy cable testers over theInternet and at some computer stores Cable testers cost between $50 and $250

Looking at twisted-pair cabling problems

Twisted-pair cabling usually displays a link light on the back of the computer, in the area ofthe network card port As long as the link light is lit, the cable doesn’t have a problem.Again, check for physical problems with the cable: cuts, crimps, coiled cable, and so on.Check the connectors If you have a cable tester, use it If all else fails, you can call a cableprofessional to test the cabling for you and replace it as necessary

Considering hub problems

If you have only one hub on your network and it fails, the entire network goes down If youhave multiple hubs on a network and one fails, only the segment serviced by that hub goesdown It’s also possible that one port on a hub will fail — and that affects only one computer.That is a difficult situation to troubleshoot, because you usually start diagnosing the PC, net-work card, and cabling before you get to the hub

If you think you have a cabling problem but find nothing wrong with the cable, change thehub port to see if that’s the solution

You also can remove one workstation at a time to see if the other computers work If ing one workstation eliminates the problem, there’s something wrong with that workstation’scable Make sure that the cable length is acceptable

remov-Examining bottlenecks

If your network is slow and everyone’s complaining about it, you probably have a bottlenecksomewhere in the system A bottleneck is usually one network component causing sluggishtraffic flow If all your networking hardware works well together, you won’t have bottlenecks

If you do, you need to locate the problem and alleviate it

Memory is always a prime suspect for a slow computer or network A server, for example,

or a workstation that carries most of the load could be a bottleneck on the network Wheneveryone accesses one computer, its processes slow down Adding memory can improve thesituation

Applications often create a bottleneck if they’re faulty or a little buggy Make sure that youuse the latest upgrades to any program installed on your computers If you think a program isrunning slowly, check with the manufacturer to see if there is an available upgrade or a patchthat solves the problem

When anyone on the network transfers several large files — such as images, sound clips, ormotion files — from one computer to another, that transfer can cause a log jam You mightsave those files to a Zip disk or CD-RW to help ease a network traffic problem

Investigating Peripheral Problems

Each peripheral in your computer can cause multiple problems by itself or in conjunction

with other peripherals If you have a problem you cannot solve, check the peripheral’s

docu-mentation and the manufacturer’s Web site for extra information

Examining network printer problems

Often, simple things cause printer problems — such as an unplugged network cable or a

prob-lem with the application Other probprob-lems can be more difficult to find If you have trouble

with your printer and you’ve tried all the following suggestions, you should check with the

manufacturer for more specific information (refer to the printer documentation or the

manu-facturer’s Web site for contact information)

General printer check

As with any network equipment, you want to physically check a printer that’s giving you

problems before you check anything else Make sure that the printer is turned on and online,

check the paper tray and path, and check all power cords and cables Always try turning the

printer off and then back on again to reinitialize it, in case some settings are corrupted If

your printer doesn’t have an on/off switch, unplug the printer for 10 seconds or so and then

plug it back in

Next, run the printer’s self-test to see if the problem is in the printer or elsewhere If the

printer passes, use Print Screen from an MS-DOS prompt to see if the computer can

communi-cate with the printer The easiest method of printing with the Print Screen button on your

keyboard is to first go to the MS-DOS prompt, and then type dir so that you have something

on the screen Press the button on your keyboard labeled Print Screen to see if the directory

on your screen prints If this second test fails, replace the cable between the printer and the

computer If this second test is successful, check the application that is printing to the printer

If the printer works from MS-DOS but not from Windows, you might have a Windows problem

Try removing the printer driver and then installing it again

If the printer still won’t print, check the server if there is one, or the computer to which the

printer is installed locally For example, if the printer’s icon in the user’s Print Folder is

grayed, or dimmed, that means the printer is offline or turned off

Next, have the user print another document — something simple from WordPad, for example

Check the print queue on the computer attached to the printer to see if the job is listed If the

job isn’t listed, check to see if the network user who is trying to print is logged on to the

net-work Also check to see if the user has permission to print, if the printer is shared, and so on

Network printer check

If the previous steps don’t solve or address your printing problem, try these network printing

troubleshooting steps

Print to a local printer If that works, print to a file and copy the file to the network printer To

do this, follow these steps:

1 Choose Start ➪ Settings ➪ Printers The Printers window appears.

2 Right-click the printer’s icon and then choose Properties The printer’s Properties

dia-log box appears

3 Choose the Details tab, as shown in Figure A-11.

Figure A-11: Set to print to a file.

4 In Print to the Following Port, choose FILE: (Creates a file on disk).

5 Click OK to close the dialog box.

6 Print any file from any application When prompted, enter a filename for the print job.

To print the file, open the MS-DOS window and, at the prompt, type the following:

copy <path\filename> <networkprinter> /b

For example, type copy c:\docs\letter.prn \\Sue\win98hp /b The /b makes the file a binary

one, which means the entire file will print If you don’t use a /b, only part of the file might

Solving Spool32 errors

When you’re printing to a network printer, you might receive a Spool32 error Spool32.exemanages print spooling, which enables the computer to process a job more quickly and effi-ciently Following are three samples of Spool32 errors:

SPOOL32 caused a Stack Fault in module Kernel32.dll at (address).

SPOOL32 caused an Invalid Page Fault in module Kernel32.dll at (address).

SPOOL32 caused a General Protection Fault in module Kernel32.dll at

(address).

Your first step should be to make sure you have the most updated printer driver for yourprinter Check the printer manufacturer’s Web site (or call the manufacturer) to see if anupdated version of the driver has been released Often, manufacturers release another version

of their printer driver when they find a coding error in the previous driver version Replacingthe driver not only clears the Spool32 error but also makes your printer run more efficiently

An antivirus program, a terminate-and-stay-resident program (a DOS program that stays

loaded in memory, such as calendars or calculators), or some other program could be

block-ing your printblock-ing You can figure out what is blockblock-ing the printblock-ing by followblock-ing these steps:

1 Disable the config.sys, autoexec.bat, winstart.bat, system.ini, and win.ini files all at the

same time You can do this by using System Information, as described in the section

“Using Microsoft System Information,” earlier in this chapter

2 Restart the computer Now print If you still have problems, go to Step 3 If this takes

care of the problem, complete your printing, and then go back to System Information

and enable the files you disabled in Step 1

You probably won’t run into a Spool32 problem very often If you do, however, you can find

out which program is causing the problem by enabling only one of the files listed in Step 1,

starting your computer, and printing If the printing works, enable another file and try again

When one file blocks the printing, open the file in a text editor and see what programs it’s

loading You can disable each program, one at a time, to find the cause of your Spool32 error

3 Print directly to the printer by changing the spooling format from Enhanced Metafile

(EMF) to RAW You do this in the Details tab of the printer’s Properties dialog box In

the Spool Settings dialog box, click Print Directly to the printer; in the Spool Data

Format box, click RAW

Looking at remote connection problems

As often as not, remote connection problems are the fault of the remote computer Before you

spend too much time on your end searching for the cause, check with the person running the

remote modem and computer to see if he or she has a problem

Printing and Dial-Up Networking

If you have trouble printing a remote page, such as a Web page or an intranet page, over a

dial-up connection, you might be trying to print on a different subnet than your dial-dial-up connection

You can work around this problem by printing the page while on the Web or intranet Next,

quit the Dial-Up Networking connection Choose Start ➪ Settings ➪ Printers Right-click the

network printer, and then click Use Printer Offline to remove the check from the check box

Press F5 to refresh the printer’s status and close the Printers folder The Web page should

print normally now

Error 645 with Dial-Up Networking

An error 645 occurs when Dial-Up Networking cannot complete a connection The error reads

as follows: “Dial-Up Networking could not complete the connection to the server.”

The problem generally occurs when the Require Encrypted Password option is enabled To

alleviate this problem, follow these steps:

1 Open My Computer.

2 Double-click Dial-Up Networking.

3 Right-click your connection icon and choose Properties.

4 Select the Server Types tab and then click the Require Encrypted Password check box

to remove the check

5 Click OK.

Tip

Problems opening Dial-Up Networking

Problems with your DLL files can result in the Dial-Up Networking folder disappearing, notopening, or appearing briefly before disappearing Even reinstalling Windows doesn’t help.The problem could be caused by any of the following DLL files being missing or damaged:

✦ Wsock32.dll

✦ Wsock32n.dll

Windows 98 doesn’t have the Wsock32n.dll file unless you have upgraded from Windows 95

or you installed a program that also installed the file

✦ Rnaui.dll

✦ Msvcrt20.dll

✦ Rasapi32.dll

✦ Rnaapp.exeYou can resolve the problem by following these steps:

1 Restart the computer, but when you see Starting Windows appear on-screen, press the

F8 key

2 From the menu, choose Safe Mode, Command Prompt Only.

3 Rename the Wsock32.dll, Wsock32n.dll, Rnaui.dll, Msvcrt20.dll, Rasapi32.dll, and

Rnaapp.exe files (found in the Windows\System folder) to the same filenames but use adifferent extension, such as aaa or ccc Wsock32.dll, for example, becomes Wsock32.aaa,and so on Be careful not to use an extension often used in Windows, such as exe or com

If the files do not exist, skip this step

4 Restart Windows.

5 Extract a new copy of the files from the original Windows CD-ROM See online help for

information on extracting files

Decreasing logon time to ISPs

Often you see a long logon delay when you’re dialing your ISP or other service provider Bydefault, Windows enables the software compression and log on to network options for Dial-

Up Networking It also binds NetBEUI, IPX/SPX, and TCP/IP to the dial-up adapter by default

By changing some of these options, you can speed up the logon process

You can cut down the delay time in Windows 98 by doing the following:

1 Open My Computer.

2 Double-click Dial-Up Networking.

3 In the Dial-Up Networking window, select the connection Right-click the connection

and choose Properties

4 In the Properties dialog box, choose Server Types Figure A-12 shows the default

options that are set

5 Deselect Log on to the Network, NetBEUI, and IPX/SPX Compatible.

6 Click OK.

Note

Figure A-12: Disable some default

options

Solving Specific Problems

You can experience some pretty specific problems on a network The Internet is a wonderful

source for finding solutions to networking problems You also might try to contact the

manu-facturer of the products you’re using or take a look at the Web site of the manumanu-facturer or

vendor for ideas Following are some of the more common problems you may encounter

✦ Computer crashes randomly and inconsistently — Insufficient RAM usually causes

fre-quent and random computer crashes You might want to check to see if the RAM chips

are well seated first If a chip is not well seated, remove and replace the RAM chip

✦ Conflicts with a card or other device — If you have problems with installing a new

adapter card or other device, check to see that the IRQ isn’t already in use IRQs are

limited in number, and no two devices can share the same one For more information,

see Chapter 25

✦ Defective data, printing problems, monitor static — Power problems generally are

related to damaged computer components Line noise, power surges, spikes, and

black-outs can overheat your hardware or corrupt data If you notice problems with your

lights or other electrical equipment, you can assume your computer equipment also is

affected by these power problems A UPS and surge protector are your best defense

See the section “Examining a UPS,” earlier in this chapter, for more information

✦ Net Watcher doesn’t disconnect the user — Sometimes Net Watcher doesn’t

discon-nect a client user when you indicate you want the user discondiscon-nected This is a minor

glitch in the program It disconnects the user, but when the user logs on to the network

again, he or she can reconnect to the resource If you don’t want a user connected to a

resource, you must change the password on the resource

✦ Sluggish system resources — Your system resources — memory, processor speed,

caching, and so on — might seem to decrease even though no programs are running

This situation happens when you start a program and then quit it before you let it start

all the way Stopping the program before it starts “leaks” memory and decreases system

resources Leaked memory consists of data that is stored in temporary RAM and takes

up space until you turn off the computer The only way to resolve this problem is torestart the computer If you still have the problem, check carefully for viruses on yoursystem Viruses often work in the background to eat system resources

✦ Problems with the System Monitor — If your System Monitor suddenly quits without

notice as you view other computers on the network, it’s probably because someone onthe network pressed Ctrl+Alt+Delete This action not only causes the networked com-puter to reboot or display the task list, but it also hangs the System Monitor

Solving share problems

If you need to look at the network resources and their shares for the network, you can use theNET VIEWcommand Use the command at the MS-DOS prompt if you’re having trouble locating

a shared resource NET VIEW is great for finding the drives, folders, printers, and other names

of resources

To view shared resources, follow these steps:

1 Choose Start ➪ Programs ➪ MS-DOS Prompt for Windows 98 or Start ➪ (All) Programs ➪

Accessories ➪ Command Prompt for Windows 2000 or XP

2 At the C prompt, type the following command:

net view \\computername

For example, type net view \\sue.

The command displays the share name for the drives, folders, printers, and other resources;the type of share; and any comments about the share Figure A-13 shows the results of thecommand example in Step 2 Note that the folders, a Zip drive, and a printer are shared onthe referenced computer

Figure A-13: Find out about the shared drives, folders,

and other resources

Missing the Network Logon dialog box

Sometimes the network logon dialog box disappears during logon If the dialog box isn’t therefor the user to enter a username and password, the user cannot access the network resources.Any of several causes might be responsible; following are the more common reasons:

✦ AutoLogon in the Registry is a number other than 0.

✦ The network interface card is improperly configured

✦ The Primary Network Logon field isn’t set correctly

To resolve the problem, check your Network dialog box first (choose Start ➪ Settings ➪ Control

Panel ➪ Network icon) in Windows 98 or, for both Windows 2000 and XP, right-click My Network

Places ➪ Properties then right-click Local Area Connection and choose Properties Check to see

if the network card configurations are correct You should check to see that the card is bound

to the correct protocol and that the IRQ and I/O addresses are correct See Chapter 25 for more

information If you made any changes, restart your computer to see if the Network Logon dialog

box comes up now

Before you close the Network dialog box, check the Primary Network Logon in the

Configuration tab, as shown in Figure A-14 The primary logon should be set to Client for

Microsoft Networks If it’s been changed to a Windows Logon, the Network Logon dialog box

will not appear If you changed the primary network logon, restart the computer

If neither of these configurations solves the problem, you need to change an entry in the

Registry The following entry will only be listed in your Registry if you’re on a client/server

network Follow these steps:

1 Choose Start ➪ Run The Run dialog box appears.

2 In the Open text box, enter regedit and then press Enter The Registry Editor appears.

3 Locate the HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\

Network\Real Mode Net key

4 Locate the AutoLogon entry It should read AutoLogon=0 If it doesn’t list 0, change the

value to 0

5 Close the Registry and restart the computer.

Figure A-14: Make sure that you’re on

the right network

Troubleshooting with the System

Configuration Utility

Windows 98 and XP have a program that helps you diagnose system configuration issues TheSystem Configuration utility is also called msconfig enables you to select check boxes thatturn off certain settings so you can find out which settings are causing the problem msconfig

is especially useful in locating viruses in your system that could affect your computer and theentire network

You can set your configuration preference for the following files in the System Configurationutility:

✦ System.ini

✦ Boot.ini

✦ Win.ini

✦ Programs that startup with the computer

✦ Environment and International settings (in Windows XP)

In XP, you must log on as an administrator to use msconfig To start msconfig, follow these steps:

1 Click Start ➪ Run The Run dialog box appears.

2 Type msconfig in the Open text box and click OK.

You can perform various tasks with msconfig On the General tab, you can choose to start thecomputer with a diagnostic startup in which Windows loads only the basic devices and services,step-by-step, so you can determine if a device or service is causing your problem When you usediagnostic startup, networking, Plug and Play, and other services are temporarily disabled.You can alternatively use the Selective Startup option on the General tab Deselect processing

of the system.ini, boot.ini, win.ini, and so on to see if one of those files is causing a problemwith Windows When you use selective startup, deselect only one process or system at a timeand then reboot If you choose multiple services or systems to disable, you won’t know whichone is causing the problem

You can stop individual files from loading on boot in the System.ini, Win.ini, or Boot.ini tabs.(The Boot.ini tab is available only in Windows XP.) Naturally, if you’re unsure of what you’redoing and you’re not sure what these files do, you shouldn’t edit the configurations; youcould disable your computer

A handy tool for locating viruses is the Startup tab of the System Configuration Utility TheStartup tab lists the programs, the program command and path, and the location of the filesthat open when you start up your computer By deselecting a startup item, you can stop yourcomputer from booting with that item You can select the item again if and when you find theitem was not causing a problem

A virus might show up in the Startup tab of your computer For example, Figure A-15 showsthe run= command at the bottom of the list in the Startup tab The alexir.exe andinstit.batare both leftover viruses The antivirus quarantined the files, but the run= com-mand in startup still searches for them when the computer boots Deselecting the line in theStartup tab stops the operating system from looking for the virus files

Figure A-15: Stop individual items from starting up with the operating system.

Troubleshooting Macs

There are many troubleshooting procedures to go through if you’re using an AppleShare

network — with all Mac computers This book, however, is about using Macs with PCs, so

this section is targeted to those networking problems you might have with your mixed

net-work If you have trouble with AppleShare or with your individual Mac computer, see the

documentation or go to www.apple.com for more information

You should use ping and other network troubleshooting tools just as you do with a PC to

make sure the hardware connection is still viable If the ping doesn’t work, then check

cables, network card, hubs or switches, and so on

Next, verify that the Mac is correctly configured for TCP/IP Many issues can cause problems

with attaching to the network: A TCP/IP setting could accidentally change, a duplicate IP

address could appear on the network, a system file or TCP/IP preference file might become

corrupted

In addition, you should check the following things on your Mac and network:

✦ Make sure network wires are strong and stable, or if wireless, make sure the network

follows the wireless distance and direction guidelines

✦ Swap out the network card with a working computer

✦ Reset the PRAM (see your Apple documentation)

✦ Delete TCP/IP settings and reconfigure

✦ Turn off non-MacOS extensions (consult your Apple documentation).

✦ Reinstall Open Transport

✦ Reinstall the system software

You can also check out some of these Wiley books for help and information about your

Macintosh: Mac OS X For Dummies, 2nd Edition; Mac OS Bible, Jaguar Edition; and Mac OS 9 For

Dummies.

Troubleshooting Linux

Linux comes in so many different flavors that it’s hard to offer specific help You can, ever, purchase books, go to online forums, and check the Web for information troubleshoot-ing your distribution Following are some interesting sites that help you troubleshootproblems with Linux:

how-✦ www.linuxquestions.org for questions, forums, FAQs, a glossary, and more tion about Linux

informa-✦ http://www.ibiblio.org/pub/Linux/docs/ for documents that answer a variety ofquestions and troubleshoot problems with Linux

✦ www.cnet.com is a good search engine for help with Linux, plus CNET offers advice andtips for Linux users

✦ www.linuxreinstall.com/debian.htm offers troubleshooting information about theDebian distribution

✦ www.linuxreinstall.com/suse.htm offers troubleshooting for the Suse distribution

of Linux

✦ www.linux-mag.com offers experts to answer your questions about Linux

In addition, following are a couple of books you can purchase to help you with specific

distri-butions of Linux Wiley publishes the Red Hat Linux 9 Bible by Christopher Negus (2003) plus many other Linux titles, including Red Hat Linux 9 For Dummies by Jon “maddog” Hall and Paul G Sery (2003), the WROX title Beginning Red Hat Linux 9 by Sandip Bhattacharya et al (2003), and Linux For Dummies, 4th Edition by Dee-Ann LeBlanc, Melanie Hoag, and Evan Blomquist (2002).

IP Addressing

Transmission Control Protocol/Internet Protocol (TCP/IP) is an

excellent protocol for using in a home or small-business network

Although it is difficult to configure, after you get the hang of it, you

will be pleased with the results TCP/IP is an efficient protocol that

speeds up your network communications Windows contains

every-thing you need to install and configure TCP/IP on your network

computers

Understanding TCP/IP

TCP/IP is actually a set, or suite, of network protocols designed by the

government’s Advanced Research Projects Agency in the 1970s Each

of the protocols that are part of TCP/IP contributes to making it an

efficient and fast communications language for the Internet and for

local and wide area networks (LANs and WANs)

TCP/IP works with a variety of hardware and software products You

can use it on Unix computers, Macintoshes, and PCs using Windows,

NetWare, OS/2, and more Many manufacturers and vendors support

TCP/IP because it is so widely used

The most common hardware connection for home or business

TCP/IP networks is Ethernet If you’re using a phone line network

or wireless, you need to check the documentation to make sure

that they can use TCP/IP

TCP/IP is especially useful today, because it enables local networks to

plug into the Internet easily, corporate intranets to become virtual

private networks (VPN) quickly, and VPNs to become extranets, thus

linking the world together in a variety of ways A VPN is a network

you set up that uses the Internet as your communications base The

VPN, for example, might connect two or more corporate LANs across

the country or across the world An extranet is a corporate network

that extends to the Internet so that vendors, manufacturers, and

cus-tomers can access corporate data

Although most of the discussion in this appendix refers to TCP/IP

and IP as it applies to the Internet, the topics also refer to the use

of TCP/IP on a home or business network The latter is just on a

Examining the basics

TCP/IP enables multiple computers to link together in a network If your network uses TCP/IP,

for example, you can link it to the Internet by using a router A router is an intelligent device

that connects network segments Routers transmit network packets from one network toanother

The Internet is, in fact, one huge network of servers and computers attached to each otherthrough a series of routers They can communicate with each other because of the TCP/IPprotocol TCP/IP hides the routers, though, so all computers on the Internet seem like one bignetwork

IP

One TCP/IP protocol, in particular, enables network packets to move data between networksegments and travel across routers — Internet Protocol (IP) IP is a routing protocol, meaning

it directs datagrams from the source to the destination Datagrams are packets of data that

contain the source and destination addresses, as well as the data Each datagram contains allthe information it needs to find the target host, no matter which paths the other datagramshave taken

IP chooses the path the packets take across routers and networks IP regulates packet warding by tracking Internet addresses, recognizing incoming messages, and routing outgoingmessages Datagrams might arrive at the destination in any order, however, or they might notarrive at all

for-TCP

Transmission Control Protocol (TCP) is a higher-level protocol than IP It provides continuingconnections between programs TCP also makes IP datagrams smaller and faster

TCP divides datagrams into smaller segments to fit the physical requirements of the servers

on the network It then uses IP to transmit the segments of data IP doesn’t guarantee the ments will arrive at the destination in the same order as they were sent

seg-TCP inserts a header into each segment that is used to track every segment from one port tothe other TCP guarantees that every byte sent arrives — and without duplication or loss.After the segments arrive at the target host, TCP checks for errors If it finds any corrupteddata, it discards that data and requests that the data be transmitted again

UDP

User Datagram Protocol (UDP) is similar to TCP in that it divides some datagrams into ments and sends them over the network by using IP UDP is, in fact, a primitive version ofTCP TCP performs error-checking tasks to make sure that the segments arrive without anyfaults or problems UDP, on the other hand, doesn’t guarantee that the datagrams will arriveintact or even at all

seg-UDP is used in many programs today For example, seg-UDP works with SNMP applications on thetransport level of communications SNMP, which is short for Simple Network ManagementProtocol, is a protocol that manages and monitors the network

Miscellaneous TCP/IP protocols

Some of the other protocols in the TCP/IP suite include the following:

✦ Point-to-Point Protocol (PPP) enables connections between hosts and networks and thenodes (routers, bridges, and so on) in between

✦ Simple Mail Transfer Protocol (SMTP) is for exchanging e-mail.

✦ File Transfer Protocol (FTP) is for transferring files FTP enables one computer to

trans-fer a file to another computer by using TCP

✦ Server Message Block (SMB) enables a computer to use network resources as if they

✦ Address Resolution Protocol (ARP) translates 32-bit IP addresses into physical network

addresses, such as 48-bit Ethernet addresses

An Ethernet address is also called a MAC address (MAC stands for Media Access Control) It’s a

number written as 12 hexadecimal digits — 0 through 9 and A through F — as in 0080001021ef

Alternatively, a MAC address might have six hexadecimal numbers separated by periods or

colons, as in 0:80:0:2:21:ef The MAC address is unique to each computer, but it does not

iden-tify the location of the computer, only the computer itself

✦ Reverse Address Resolution Protocol (RARP) translates physical network addresses

into IP addresses

✦ Internet Control Message Protocol (ICMP) helps IP communicate error information

about the IP transmissions

✦ Internet Group Management Protocol (IGMP) enables IP datagrams to be broadcast to

computers that belong to groups

Some of the TCP/IP protocols are also applications FTP, Telnet, and SNMP, for example, are

programs that you can use over the network because they’re included with the TCP/IP suite

Glimpsing the ISO/OSI model

International Organization for Standardization/Open Systems Interconnection (ISO/OSI) is a

set of standards that defines network functionality ISO/OSI sets standards for cabling,

net-work interface cards (NICs), protocols, and so on

TCP/IP’s layered design works well with the ISO/OSI model to transmit network data efficiently

and effectively As the data moves from the network application (Layer 7) to the network card

(Layer 2), one or more of the TCP/IP protocols accompanies it every step of the way

The seven-layer model defines computer-to-computer communications Following is a brief

explanation of each layer:

✦ Layer 1 — The physical layer defines the cabling.

✦ Layer 2 — The data link layer controls the flow of data through the network cards.

✦ Layer 3 — The network layer defines the protocols for data routing, to make sure that

the data gets to the correct destination

✦ Layer 4 — The transport layer defines protocols for error checking and message

formation

Note

Note

✦ Layer 5 — The session layer maintains the connection, or session, for as long as it takes

to transmit the packets The session layer also performs security and administrationfunctions

✦ Layer 6 — The presentation layer identifies the way the data is formatted.

✦ Layer 7 — The application layer defines how the applications interact with the network.

Exploring IP Addressing

IP addressing was standardized in 1981, with specifications that required each systemattached to the Internet be assigned a unique, 32-bit address value Systems include servers,routers, gateways, and other networking hardware A router that attaches two network seg-ments, for example, must have two unique IP addresses, one for each network interface

To ensure that IP addresses used on the Internet are unique, the Internet Network InformationCenter (InterNIC) must assign any address used on the Internet InterNIC is the controllingagency for IP addresses and domain names

Examining an IP address

An IP address identifies the computer or other node (router, printer, server, or other) on the

network Each IP address on a network must be unique

An IP address is a binary number written in a series of four decimal digits, which is known as

dotted decimal Four period-delimited octets consisting of up to 12 numerals forms an IP

address For example, Microsoft’s home page IP address is 207.46.131.137 and is a dotted imal The numbers represent decimal notations for each of the four bytes of the address; theaddress identifies the computer

dec-The IP address is really made up of two parts: the network number and the host number

✦ The network number identifies the general location of the computer on the network,

and the host number pins it down to the exact computer In Microsoft’s IP address,207.46 is the network address

✦ The host number is represented by 131.137 Each class of address uses a different

man-ner of dividing the octets Microsoft is a Class B network (see the following section).The highest value in any octet is 255, because of the way the binary format translates to dot-ted decimal format

Understanding address classes

IP addressing is divided into five categories, or classes Three of the classes — Class A, Class

B, and Class C — are in use today The following list describes each of the classes:

✦ Class A is used for large networks To identify a Class A network address, the first octet

uses the numbers from 1 to 126 Class A networks have an 8-bit network prefix; fore, they are currently referred to as /8s (pronounced “slash eights”) or just “eights.”

there-✦ Class B is mainly used for medium-sized networks, and the first octet values range from

128 to 191 Class B network addresses have a 16-bit network prefix; thus, they arereferred to as /16s

Note

✦ Class C is reserved for smaller networks To identify a Class C network, the values

range from 192 to 233 Class C networks have a 24-bit network prefix, and so are

All Class A addresses already have been taken by universities and corporations Class B

addresses are assigned to companies and institutions with a minimum of 4,000 hosts If you

apply for an Internet address, you will probably receive a Class C designation

Each class defines its own 32-bit address boundaries In Class C, the first three octets are for

the network address; the last octet represents the host address If you apply for an Internet

address, the InterNIC will give you an address with the first three octets defined You fill in

the last octet with numbers ranging from 1 to 254 The numbers 0 and 255 are reserved Each

number you assign goes to one node on your network, so you can connect up to 254 nodes to

the Internet

Looking at the subnet mask

A subnet mask is part of the IP addressing system A subnet mask creates subnetworks that

enable a computer in one network segment to communicate with a computer in another

seg-ment of the network The main reason for subnetting (or creating subnets on a network) is to

divide a single Class A, B, or C network into smaller pieces

The subnet mask is a 32-bit address that hides, or masks, part of the IP address so as to add

to the number of computers added to the network All networks must use a subnet mask,

even if they don’t connect to another network If a network isn’t divided into subnets, the

default subnet mask is used The default depends on the IP address class

✦ Class A networks use a default subnet mask of 255.0.0.0

✦ Class B uses a default subnet mask of 255.255.0.0

✦ Class C uses a default subnet mask of 255.255.255.0

Subnetting enables organizations to mix different network technologies across several

physi-cal segments It also enables you to exceed the maximum number of hosts per segment, if

you’ve used all your IP addresses

Comprehending the gateway

The gateway is a bridge between two segments of a network Messages travel between

net-work segments through the gateway A gateway is a combination of hardware and software; it

creates a shared connection between, say, a LAN and a larger network

Often, you use a gateway to bridge two networks that use different communications

proto-cols A gateway has its own processor and memory that it uses to convert protocols;

convert-ing protocols makes the gateway slower than a router or bridge A gateway must have its own

IP address

Working with Domain Names

Every IP address on the Internet has a corresponding domain name, such as microsoft.com.Domain names make it easy to remember addresses, and you can use them in place of the IPaddress in the URL text box of your browser

Regarding domain names

IP addresses are difficult to remember, so domain names also represent a computer on theInternet Microsoft’s domain name, for example, is www.microsoft.com Domain names usu-ally start with www, which stands for World Wide Web; however, www is not always included

in an address The letters www represent a route to a World Wide Web server Other servers

might use a different route, such as w2 If no route is listed, the address takes a default route

as listed by the server

Domain names are written from the least specific (top-level domain) to the most specific(host name) Each part is separated by a dot ( )

The second part of the domain name is the name of the organization, company, product, oranother catchy word or phrase Microsoft, for example, is perfect for the domain name oftheir Web site

The third part of the domain name identifies the type of organization The letters com, for

example, stand for commercial Following are other top-level domain identifiers and theirmeanings:

Other additions to the domain name include a country code, if the server is located outside

of the United States UK stands for United Kingdom, for example, and IT stands for Italy.Domain names are listed in the Uniform Resource Locator (URL) to a site The URL is the fulladdress, or computer identifier, on the Web URLs contain numerous slashes and dots thatseparate the parts of the address, similarly to the way you separate folders in a path.Microsoft’s complete URL is http://www.microsoft.com/ If you want a particular docu-ment on a site, you must use a longer URL, such as http://support.microsoft.com/sup-port/index.html

The letters http stand for Hypertext Transfer Protocol; this is the protocol your computer

uses to attach to the server computer HTTP defines the language the computers will use totransfer pages and hypertext (links) With most newer browsers, you don’t have to type thehttp in the address, but it doesn’t hurt to add it

In the sample URL, no www is used, but the address of the server is identified as support.microsoft.com The support between forward slashes represents a folder, or directory

Tip

The HTML document is the one you view in that folder There are no rules as to when to use

wwwor http The best practice is to copy the exact URL from literature or documentation

about the Web site

Understanding the domain name system

The Domain Name System (DNS) is a method of matching IP addresses with domain names.

When you type a domain name in the URL address area of your browser, that query is

trans-mitted to a DNS server A DNS server maintains a database of domain names and IP addresses

The DNS server finds the IP address that matches the domain name and then sends your

request on to that server The process is called name resolution.

You might find a DNS server in a university or college, on a corporate LAN, or even on a

smaller LAN Most primary ISPs also have DNS servers Local ISPs connect to larger, or

sec-ondary, ISPs, and those ISPs connect to much larger, primary ISPs that make up the Internet

DNS, or name, servers are grouped into domains, which identify different levels of authority

At the top of this hierarchical structure is the root domain, or top-level domain, such as com,

edu, org, and so on

Within the top-level domains are second-level domains Second-level domains contain hosts

and subdomains Going back to the microsoft.com example, microsoft.com is a second-level

domain A subdomain of microsoft.com might be ftp.microsoft.com

support.microsoft.comthen would be a host name within the domain

Each Domain Name Server has a specific area for which it stores addresses and domain

names Called the zone of authority, the Domain Name Server can resolve only addresses

within its zone If a Domain Name Server doesn’t contain the IP address for the queried

domain name, it forwards the query to another Domain Name Server

Recognizing DHCP

Dynamic Host Configuration Protocol (DHCP) is a utility for assigning TCP/IP addresses to

work-stations automatically DHCP servers are used mainly by corporations and large TCP/IP networks

to configure their clients, although they are also used by some Internet service providers

When a client accesses a TCP/IP network, the DHCP server assigns the client an IP address, a

subnet mask, and a gateway, if needed The DHCP server has a range of possible IP addresses

from which to choose Each time the client logs off of the network, the IP address goes back into

a pool and might be assigned to another client logging on to the network Letting the DHCP

server configure clients when they attach to the network saves the administrator of a large IP

net-work a lot of time and effort

A DHCP server can grant a lease to a DHCP client The lease provides one IP address for use by

the client for up to 30 days Each time the client logs on to the IP network, the lease is updated,

so the lease doesn’t run out unless the client doesn’t log on within the 30 days of the lease A

lease is important for some networks, because to communicate with other corporate programs

over the network, the client might need a stable IP address

The DHCP protocol requires a client ID for each computer By default, the DHCP server uses the

client computer’s MAC address

DHCP works with Mac OS; Microsoft Windows 95, 98, NT, 2000, XP; Novell NetWare; Novell LAN

Workplace for DOS (for attaching to a Unix workbox, or terminal); and Linux

Why Use TCP/IP?

You can use TCP/IP to connect your LAN to the Internet or to create an intranet or simply tospeed up your local area network You might want to use TCP/IP because it’s so compatible withmany different hardware and software products, or simply because you want to experiment with

it In addition, most newer operating systems — Windows 98 Second Edition, 2000, and XP; Mac

OS X 10; and Linux distributions — come with TCP/IP installed Many OSs use only TCP/IP.Although TCP/IP seems severe and awkward, it’s really not that difficult to get the hang of it.You might start by configuring your computers for dialing up the Internet and then decideyou want to try a TCP/IP network in your own home

For information about intranets, see Chapter 19 For information about using the Internet,see Chapter 16 For instructions on setting up dial-up networking, see Appendix C

Looking at advantages of TCP/IP

Using TCP/IP has many advantages It’s more efficient and faster than NetBEUI Additionally,many hardware and software vendors support TCP/IP You can use it with a variety of pro-grams and products If the computers on your network use different operating systems, such

as Macintosh and PCs, TCP/IP enables those computers to communicate

The following sections discuss some more advantages of using TCP/IP

Intranet

Using TCP/IP, you can create a private internet (intranet) on your home network Publishdocuments for your family to view on their Web browsers by using Internet technologies,including HTTP, HTML, and more

If you create an intranet for your employees to contribute to and use, you can easily connect

it to the Internet and extend its purpose to clients and vendors, as well

Following are some other reasons to create an intranet on your home network:

✦ Create an intranet so that you can share family stories, pictures, and play games overthe network

✦ Familiarize yourself and your family with Internet technology and the use of TCP/IP,

as well

✦ If you’re thinking about creating a Web page for the Internet, experiment with designs

on your intranet before you publish to the Web

✦ If you’re thinking about starting a small home business, you can use your intranet toset up a home page and an index, create links to your pages, and test the site beforeyou put it up on the Internet

✦ If you’re planning to start an Internet business, set up the e-commerce site on yourintranet before taking it to the Web

Internet

You need TCP/IP to connect to the Internet Realize that there is a difference betweeninstalling TCP/IP on your network and installing it for dial-up networking use; these are twoseparate installations and uses for the protocol

Small

Business

Tip Cross-

Reference

You install TCP/IP to use with the dial-up networking feature This installation enables you to

connect to the Internet by means of a modem The TCP/IP communicates with Web servers

on the Internet If you install TCP/IP on your network, you use it to connect to the other

com-puters on your LAN One installation has nothing to do with the other For more information,

see the section “Installing and Configuring TCP/IP” later in this appendix

Using TCP/IP to connect to the Internet provides you with all the Internet features and

advan-tages, including e-mail, Web browsers, and file transfer Also, TCP/IP includes various

pro-grams (which are also protocols) that you can use to access information on the Internet If

you have special needs for FTP or Telnet, for example, TCP/IP provides those programs for

you to use

If you plan to expand your network to the Internet someday, your knowledge and use of

TCP/IP will help you tremendously Knowing how to configure TCP/IP will make it easier for

you to transfer your Web sites and pages to the Web and to configure clients and servers for

the Internet Having TCP/IP in place means less configuring when you move to the Internet

Dial-Up Networking

You also can use TCP/IP to connect to your home or work computer If you’re on the road, for

example, and want to call your home network to retrieve some information or your schedule,

you can use Windows dial-up networking to accomplish this

You use TCP/IP in Terminal Services as well Terminal Services is a method of communication

between a remote computer and a server using Windows 2000 Server For more information,

see Appendix C

If your work network uses TCP/IP, you can call in to that network from home or while you’re

on the road Additionally, if your office network includes DHCP server, calling from a remote

computer becomes even easier The DHCP server configures your computer so that all you

need to do is dial and connect

Creating your own TCP/IP network

When setting up your own TCP/IP network, you need to choose IP addresses for your

com-puters Remember, each computer on the network needs a unique IP address You also need

to set a specific subnet mask for your network to use

Several IP addresses are reserved for private use Following are the three blocks reserved for

IP addresses:

10.0.0.0 to 10.255.255.255

172.16.0.0 to 172.31.255.255

192.168.0.0 to 192.168.255.255

For your home network, for example, you could use the following IP addresses for five

com-puters on the network:

Alternatively, you could use the following for your computers:

192.168.0.100192.168.0.101192.168.0.102192.168.0.103192.168.0.104You can change numbers only in the last octet of the IP address for a home network or small-business network If your corporate network is very large, you can make other changes to the

IP addresses, as long as they are consistent

In addition to IP addresses, you need a subnet mask Use the same subnet mask for all puters on the network The subnet mask 255.255.255.0 works very well

com-Using Windows 98 LinkLocal

Windows 98 includes another method of configuring a TCP/IP network called LinkLocal.LinkLocal creates automatic private IP addressing You can use LinkLocal if you have a networkthat doesn’t use a DHCP server and if your computer is not a host computer on the Internet.With LinkLocal, your Windows 98 computer can assign itself a private IP address withoutyour intervention LinkLocal IP addresses always begin with 169.254 Windows fills in the

other two octets for you LinkLocal uses a block, or range, of IP addresses, from which it

assigns the addresses of computers on your network Using LinkLocal makes administeringthe network much easier than filling in the TCP/IP information yourself

Following are some rules and guidelines for using LinkLocal:

✦ LinkLocal does not work on your network unless all the computers are runningWindows 98

✦ If your computer is a desktop or tower, it can assign itself an automatic private IPaddress when it starts up — as long as no DHCP server is found and the computer doesnot have a valid DHCP lease

✦ If you connect to a DHCP server later, the DHCP server-assigned IP address takes dence over the automatic private IP address

prece-✦ If the computer is a laptop, it can assign itself an automatic private IP address as long

as no DHCP server is present on the network

✦ If your computer is using a DHCP service and that service fails for some reason, thecomputer assigns itself an automatic private IP address to use until the DHCP service isback up and working It then surrenders the IP address to the DHCP server

To enable LinkLocal, follow these steps:

1 Choose Start ➪ Settings ➪ Control Panel The Control Panel window appears.

2 Double-click the Network icon The Network dialog box appears, as shown in Figure B-1.

3 Select the TCP/IP Ethernet adapter and click the Properties button The TCP/IP Properties

dialog box appears For information about installing protocols, see Chapter 10

4 In the IP Address tab of the TCP/IP Properties dialog box, select the Obtain an IP

Address Automatically option, as shown in Figure B-2

Figure B-1: Set the LAN TCP/IP in the

Network dialog box

5 Click OK to close the TCP/IP Properties dialog box.

6 Click OK again to close the Network dialog box Windows prompts you to restart the

computer

When you start the computer again, it automatically assigns itself an IP address You must

repeat these steps with each computer on the network to complete the process When you’re

finished, your network will be up and running using the TCP/IP protocol

Figure B-2: Set the option to automatic

IP addressing

Installing and Configuring TCP/IP

You can install TCP/IP on your Windows computer without adding any other software orhardware Windows 98, 2000, and XP come with Microsoft’s TCP/IP protocol suite built in; allyou need to do is configure it It’s important to note that you can install TCP/IP for two differ-ent purposes — LAN use or dial-up use

For LAN use, install TCP/IP on the Network dialog box and bind that protocol to a networkinterface card, or adapter Binding assigns the protocol to the device so that they can worktogether for communications over the network

For Internet or remote access use, install TCP/IP through the Dial-Up Networking window Youinstall the protocol on a specific connection

See Chapter 10 for more information about using TCP/IP on the LAN See Appendix C forinformation about using TCP/IP for remote access If you want to use TCP/IP to connect tothe Internet, see Chapter 16 for more information

Using TCP/IP with Windows 98

When you install TCP/IP on a computer on the LAN, you enable it to communicate with othercomputers that use the same protocol You must install TCP/IP on each computer on the net-work, use a unique IP address for each computer, and use the same subnet mask for eachcomputer you configure

You do not use this method of installation if you used LinkLocal to activate automatic private

Web Servers and TCP/IP Addressing

Using private IP addresses doesn’t affect your use of the Internet You still can attach to theInternet by using your dial-up networking TCP/IP You can add a Web page to your ISP’s Webserver, if you want Your LAN IP addresses are separate from your Internet connection

Adding a Web server to your network, however, might complicate matters a bit (A Web server is

a computer you attach to the Web so that you can display your own Web site.)

Generally, you install a special operating system (such as NT Server) and a Web server application(such as Internet Information Server) on the computer You also should install a proxy server orfirewall to keep Internet users from accessing your LAN through the server Proxy servers and fire-walls keep people from hacking into your LAN; they can also keep your LAN users from access-ing certain Internet sites

To attach your Web server to the Internet, you need an IP address and a domain name That IPaddress applies only to your Web server, not to the rest of your network You can keep your orig-inal IP addresses for your LAN, as long as you use a firewall or proxy server to separate them fromthe Web server and none of the computers using a private IP address try to get onto the Internet

If you have another protocol installed, such as NetBEUI or IPX/SPX, you should remove it

before installing TCP/IP Although you can have two or more protocols installed, such a setup

slows down your system and makes the network less efficient

Installing TCP/IP

When you install the protocol, it automatically binds to your network card To install and

con-figure TCP/IP for use on the LAN, follow these steps:

1 Choose Start ➪ Settings ➪ Control Panel The Control Panel opens.

2 Double-click the Network icon The Network dialog box appears.

3 Click the Add button The Select Network Component Type dialog box appears, as

shown in Figure B-3

Figure B-3: Add a network component.

4 In the list of network components to add, select Protocol and then click the Add

but-ton The Select Network Protocol dialog box appears, as shown in Figure B-4

Figure B-4: Add a protocol.

5 In the Manufacturers list, select Microsoft.

6 In the Network Protocols list, select TCP/IP Click OK Windows returns to the Network

dialog box, as shown in Figure B-5 Don’t close the Network dialog box yet Continue to

the next set of steps

When you install a network adapter card, Windows might install a dial-up adapter A dial-up

adapteris a device driver that is bound to the Client for Microsoft Networks and acts as aninterface between the modem and your computer You need the dial-up adapter if you plan

to connect to another network over a phone line by using a modem For more informationabout the dial-up adapter, see the section “Comprehending the dial-up adapter TCP/IPSettings” later in this appendix

Figure B-5: TCP/IP is installed.

Configuring TCP/IP

You must configure the protocol before you can use it To configure TCP/IP, follow thesesteps:

1 Select the TCP/IP protocol that is bound to your network card (as opposed to the

TCP/IP bound to the dial-up adapter) Click the Properties button The TCP/IPProperties dialog box appears, as shown in Figure B-6

2 Select the option Specify an IP Address.

3 In the IP Address box, enter the address you’re using — for example, 172.16.1.1.

4 In the Subnet Mask box, enter the four-octet number for the subnet, such as 255.255.255.0.

5 Click OK to close the dialog box.

6 Click OK to close the Network dialog box When Windows prompts you to restart your

computer, do so

When the computer restarts, the TCP/IP is active Make sure that you install TCP/IP on all puters on the network Use a unique IP address for each computer but the same subnet mask

com-Note

Figure B-6: Configure the IP address.

The TCP/IP Properties dialog box contains other tabs in addition to the IP Address tab Each

tab offers additional configurations for TCP/IP For a home network, you don’t need to

config-ure the information on any of the other tabs; however, the following sections describe each

tab so that you can understand more about TCP/IP configuration in Windows

DNS Configuration tab

Use the DNS Configuration tab to enter a host and domain name for your network server on a

client/server network Many small networks use an Internet DNS server to resolve computer

names Most ISPs will maintain domain information for you, for a fee

Alternatively, you can maintain your own DNS server with or without the use of the Internet

If you want to use your own DNS server on the Internet, you must file with the InterNIC for a

domain name and IP address for at least two DNS servers

If you set up your own DNS server on your local LAN, you can use only one server if you

want Microsoft recommends you use two DNS servers, however, so that one can act as a

backup in case the other one fails

Figure B-7 illustrates the DNS Configuration tab with a corporation’s own DNS servers listed

Note that you can set a search order if you use more than one server as a domain naming

service

Gateway tab

The Gateway tab provides a space for you to enter the IP address of any gateway on your

com-puter The gateway acts like a bridge to connect two networks that use different protocols

Figure B-8 shows the Gateway tab of the TCP/IP Properties dialog box

WINS Configuration tab

Windows Internet Name Service (WINS) is another method of resolving computer names to IP

addresses WINS supplies a database that maintains IP addresses and NetBIOS computer

names NetBIOS is a protocol that contains commands for transmitting information from

com-puter to comcom-puter WINS must be installed on a server comcom-puter to perform name resolution

Figure B-7: Specify the DNS server

addresses

Figure B-8: Use this tab only if you

have a gateway on your network

WINS eliminates the need to broadcast computer names across the network when one puter is trying to connect to another Instead, the request for a specific computer name goes

com-to the WINS server It resolves the name com-to an IP address and immediately sends the data com-tothe target destination WINS cuts traffic on a TCP/IP network

If you have a few computers (10, 20, 40, or so), you don’t need to set up a WINS server.Network traffic won’t be affected much for so few computers If your network consists ofhundreds of computers on a TCP/IP network, however, WINS can help resolve the name/IPaddress issue quickly and efficiently

Tip

Figure B-9 shows the WINS Configuration tab After you enable WINS resolution, you enter the

IP address of the WINS server

Figure B-9: Use WINS for name resolution.

Bindings tab

The Bindings tab shows which network client is bound to TCP/IP If you’re using a Microsoft

network, TCP/IP is bound to the Client for Microsoft Networks, as shown in Figure B-10

Figure B-10: Check your bindings.

Advanced tab

The Advanced tab contains any settings specific to the protocol The TCP/IP protocol doesn’t

usually contain any advanced settings; however, other protocols do

NetBIOS tab

When you use TCP/IP, Windows automatically configures support for using NetBIOS tions over the TCP/IP protocol NetBIOS facilitates applications in transferring data over thenetwork This tab shows that NetBIOS is enabled for TCP/IP (see Figure B-11)

applica-Figure B-11: NetBIOS support is

always enabled for TCP/IP

Using TCP/IP for Dial-Up Networking and Windows 98

When you want to use your modem to dial the Internet or to connect to a computer onanother TCP/IP network, you configure TCP/IP for dial-up networking The TCP/IP configura-tion takes place in the Dial-Up Networking window and is specific to the connection

If you haven’t created a connection for the Internet or remote office, see Appendix C for mation about how to create connections

infor-To configure TCP/IP for dial-up networking, follow these steps:

1 Open My Computer.

2 Double-click the Dial-Up Networking icon The Dial-Up Networking window appears, as

shown in Figure B-12

3 Right-click the connection you want to configure — in this case, Office — and then

choose Properties from the pop-up menu The connection’s Properties dialog boxappears, as shown in Figure B-13

4 Choose the Server Types tab, shown in Figure B-14 Note that all three network

proto-cols are selected by default

Figure B-12: Open the Dial-Up Networking window.

Figure B-13: Set the properties for

the remote connection

Figure B-14: Look for the protocols area.

For information about the other options in the Server Types tab, see Appendix C

5 In the Allowed network protocols area, make sure that TCP/IP is selected Deselect any

protocol you won’t be using with this connection, such as NetBEUI and IPX/SPX

6 Click the TCP/IP Settings button The TCP/IP Settings dialog box appears, as shown in

Figure B-15

Figure B-15: Configure TCP/IP settings for

dial-up networking

Tip

7 Choose the options and enter the IP addresses for the connection, as explained in

Table B-1

8 Click OK to close the TCP/IP Settings dialog box Click OK again to close the

connec-tion’s Properties dialog box

Table B-1: TCP/IP Settings for Dial-Up Networking

Server assigned IP address Use this option if you attach to a DHCP server

Specify an IP address Use this option if you want to specify a static IP address

connection attaches

Server assigned name server addresses Choose this option if the server to which you’re attaching

assigns DNS server IP addresses

Specify name server addresses Choose this option if you want to enter the DNS or WINS

server IP addresses manually

Use IP header compression Check this box if you want to compress the headers for your

network packets sent to this computer Using compressionspeeds up transmissions

Use default gateway on remote network Check this option if a gateway separates your computer from

the computer to which you want to connect

Comprehending the dial-up adapter TCP/IP settings

You are probably wondering why you didn’t configure the TCP/IP for dial-up in the Network

dialog box (in the Control Panel) The Network dialog box contains a dial-up adapter that is

bound to TCP/IP and the Microsoft Client for Networks As mentioned, the adapter is a device

driver that enables your modem to attach to other computers via the telephone lines

If you select the dial-up adapter in the Network dialog box and then click the Properties

but-ton, you’ll see the warning dialog box shown in Figure B-16

Figure B-16: Windows warns you away from the

dial-up adapter TCP/IP settings