Network+ 2005 In Depth (P25) docx

If you must backlevel a network operating system upgrade, you should alsoconsult with experienced professionals about the best approach for your network environment.Table 15-1 Reversing

reverse an upgrade If you must backlevel a network operating system upgrade, you should alsoconsult with experienced professionals about the best approach for your network environment.

Table 15-1 Reversing a software upgrade

Type of Upgrade Options for Reversing

Operating system patch Use the patch’s automatic uninstall utility.

Client software upgrade Use the upgrade’s automatic uninstall utility or reinstall the previous

version of the client on top of the upgrade.

Shared application upgrade Use the application’s automatic uninstall utility or maintain a complete

copy of the previous installation of the application and reinstall it over the upgrade.

Operating system upgrade Prior to the upgrade, make a complete backup of the system; to

back-level, restore entire system from the backup; uninstall an operating tem upgrade only as a last resort.

sys-Hardware and Physical Plant Changes

Hardware and physical plant changes may be required when a network component fails or functions, but more often they are performed as part of an upgrade to increase capacity, improveperformance, or add functionality to the network In this section, you will learn about the sim-plest and most popular form of hardware change—adding more of what you already use, such

mal-as adding four more switches to the backbone or adding 10 new networked printers You alsolearn about more complex hardware changes, such as replacing the entire network backbonewith a more robust system

Many of the same issues apply to hardware changes as apply to software changes In lar, proper planning is the key to a successful upgrade When considering a change to your net-work hardware, use the following steps as a guide:

particu-1. Determine whether the change is necessary

2. Research the upgrade’s potential effects on other devices, functions, and users

3. If you decide to implement the change, notify system administrators, help desk personnel,and users, and schedule it during off-hours (unless it is an emergency)

4. If possible, back up the current hardware’s configuration Most hardware (for example,routers, switches, and servers) has a configuration that you can easily copy to a disk

In other cases (for example, networked printers), you may have to print the hardware’sconfiguration

5. Prevent users from accessing the system or the part of the system that you are changing

6. Keep the installation instructions and hardware documentation handy.

7. Implement the change

8. Test the hardware fully after the change, preferably putting a higher load on the

device than it would incur during normal use in your organization Note any tended or unanticipated consequences of the change

unin-9. If the change was successful, re-enable access to the device If it was unsuccessful, late the device or reinsert the old device, if possible

iso-10. Inform system administrators, help desk personnel, and users when the change is

complete If it was not successful, explain why

11. Record your change in the change management system

Adding or Upgrading Equipment

The difficulty involved in adding or upgrading hardware on your network depends largely onwhether you have used the hardware in the past For instance, if your organization always usesCisco switches, adding one more Cisco switch to your second-floor telecommunications closetmay take only a few minutes and cause absolutely no disruption of service to your users Onthe other hand, even if your company uses Cisco switches, adding a Cisco VPN router to yournetwork may be an entirely new experience You should research, evaluate, and test any unfa-miliar piece of equipment that you intend to add or upgrade on your network, even if it ismanufactured by a vendor that supplies much of your other hardware

With the rapid changes in the hardware industry, you may not be able to purchase identicalhardware even from one quarter to the next If consistency is a concern—for example, if yourtechnical staff is familiar with only one brand and model of printer, and you do not have thetime or money to retrain personnel—you would be wise to purchase as much hardware as pos-sible in a single order If this approach is not feasible, purchase equipment from vendors withfamiliar products and solid reputations

Each type of device that you add or upgrade on the network will have different preparation andimplementation requirements Knowing exactly how to handle the changes requires not only aclose reading of the manufacturer’s instructions, but also some experience with the type of net-working equipment at hand The following list provides a very general overview of how youmight approach adding or upgrading devices on the network, from the least disruptive to themost complex types of equipment The devices at the bottom of the list are not only the mostdisruptive and complex to add or upgrade, but also the most difficult to remove or backlevel

◆ Networked workstation—A networked workstation is perhaps the simplest device to

add It directly affects only a few users, and does not alter network access for anyoneelse If your organization has a standard networked workstation configuration (forexample, a disk image—a compressed snapshot of the workstation’s contents—onthe server), adding a networked workstation will be a quick operation as well Youcan successfully add a networked workstation without notifying users or support

staff and without worrying about downtime

Chapter 15 693

HARDWARE AND PHYSICAL PLANT CHANGES

◆ Networked printer—A networked printer is easy to add to your network, too Adding this

equipment is slightly more complex than adding a networked workstation, however,because of its unique configuration process and because it is shared Although it affectsmultiple users, a networked printer does not typically perform a mission-critical function

in an organization, so the length of time required to install one does not usually affectproductivity Thus, although you should notify the affected users of a networked printeraddition, you do not need to notify all users and support staff Likewise, you do not need

to restrict access to the network or worry about downtime in this instance

◆ Hub or access point—A single hub or access point might service as few as one or as

many as 64 users You do not have to worry about downtime or notifying users when

adding a new hub or access point, because it cannot affect anyone until it is actually

in use However, if you are upgrading or swapping out an existing hub or accesspoint, you must notify the affected users, because the upgrade or swap will createdowntime In addition, you must consider the traffic and addressing implications ofadding or upgrading a hub or access point For example, if you need to expand thecapacity of a TCP/IP-based network segment from 24 users to 60 users, you caneasily enough swap your 24-port hub with a 64-port hub But before doing so, makesure that the segment has been allotted enough free IP addresses to service 60 users;otherwise, these users will not be able to access the network

◆ Server—A server addition or upgrade can be tricky Typically, this type of change (unless

it is the replacement of a minor component) requires a great deal of foresight and ning Before installing a new server, you need to consider the hardware and connectivityimplications of the change, as well as issues relating to the NOS Even if you are adding aserver that will not be used immediately, you still need to plan for its installation It’spreferable to add the server while network traffic is low or nonexistent You should alsorestrict access to the new server; otherwise, one of your users could find the server whilebrowsing the network and try to save files to it or run an application from it

plan-Upgrading the hardware (such as a NIC or memory) on an existing server mayrequire nearly as much planning as adding an entirely new server Schedule upgrades

to an existing server for off-hours, so that you can shut down the server withoutinconveniencing any users who rely on it

◆ Switches and routers—Changing or adding switches or routers to a network design is

more complicated for several reasons First, this type of change can be physically ruptive—for example, it might require the installation of new racks or other supportframes in your telecommunications room Second, switches and routers usually affectmany users—and might affect all users—on the network For instance, if you mustreplace the Internet gateway for your organization’s headquarters, you will cut everyuser’s access to the Internet in the process (unless you have redundant gateways,which is the optimal setup if you rely on the Internet for mission-critical services).You should notify all users on the network about the impending change, even if youdon’t think they will be affected—sometimes a router or switch may have unin-

dis-tended effects on segments of the network other than the one it services In tion, you should plan at least weeks in advance for switch or router changes and

addi-expect at least several hours of downtime Because switches and routers are sive, take extraordinary care when handling and configuring the equipment Also,because switches and routers serve different purposes, you should rely on the manu-facturer’s documentation to guide you through the installation process

expen-Chapter 15 695

HARDWARE AND PHYSICAL PLANT CHANGES

Bear in mind that adding a new processor to a server, a new NIC to a router, or morememory to a printer may affect your service or warranty agreement with the manufac-turer Before purchasing any components to add or replace in your network devices,

check your agreement for stipulations that might apply You may be allowed to add

only components made by the same manufacturer, or risk losing all support from thatmanufacturer

CAUTION

Above all, keep safety in mind when you upgrade or install hardware on a network Never ker with the insides of a device that is turned on Make sure that all cords and devices are stowedsafely out of the way and cannot cause trips or falls Avoid wearing jewelry, scarves, or veryloose clothing when you work on equipment; if you have long hair, tie it back Not only willyou prevent injury this way, but you will also be less distracted By removing metal jewelry, youmay prevent damage to the equipment caused by a short if the metal touches a circuit If theequipment is heavy (such as a large switch or server), do not try to lift it by yourself Finally,

tin-to protect the equipment from damage, follow the manufacturer’s temperature, ventilation, static, and moisture guidelines

anti-Cabling Upgrades

Cabling upgrades (unless they involve the replacement of a single faulty patch cable) mayrequire significant planning and time to implement, depending on the size of your network.Remember from Chapter 12 that troubleshooting cabling problems may be difficult becausethe cable layout may be undocumented and poorly planned, particularly if it was installed yearsbefore and survived intact despite building changes and network growth For the same reason,

an enterprise-wide cabling upgrade is complex The best way to ensure that future upgrades go

smoothly is to carefully document the existing cable before making any upgrades If this

assess-ment is not possible, you may have to compile your docuassess-mentation as you upgrade the ing cabling

exist-Because a change of this magnitude affects all users on the network, you should upgrade the networkcabling in phases Perhaps you can schedule an upgrade of the first-floor east wing of your buildingone weekend, then the first-floor west wing of your building the next, and so on Weigh the impor-tance of the upgrade against its potential for disruption For example, if the Payroll Department is

processing end-of-month checks and having no difficulties other than somewhat slow responsetime, it is not critical to take away its access to install CAT 6 wiring On the other hand, if thebuilding maintenance staff needs a 1-Gbps connection to run a new HVAC control system, youwill probably make it a priority to take down this access temporarily and replace the wiring In thiscase, not only must you replace the wiring, but you may also need to replace hubs and NICs.For the most part, only organizations that run very small networks are able to upgrade or install theirown network cabling Most other organizations rely on contractors who specialize in this service.Nevertheless, as a networking professional you should know how to run a cable across a room, eitherunder a raised floor or through a ceiling plenum, in order to connect a device to the network.

Backbone Upgrades

The most comprehensive and complex upgrade involving network hardware is a backboneupgrade Recall that the network backbone represents the main conduit for data on LANs andWANs, connecting major routers, servers, and switches A backbone upgrade requires not only

a great deal of planning, but also the efforts of several personnel (and possibly contractors) and

a significant investment You may upgrade parts of the backbone—a NIC in a router or a tion of cabling, for example—at any time, but upgrading the entire backbone changes the wholenetwork

sec-Examples of backbone upgrades include migrating from Token Ring to Ethernet, migratingfrom a slower technology to a faster one, and replacing routers with switches (to make use ofVLANs, for example) Such upgrades may satisfy a variety of needs: a need for faster through-put, a physical move or renovation, a more reliable network, greater security, more consistent stan-dards, support of a new application, or greater cost-effectiveness For example, the need for fasterthroughput may prompt an upgrade from an older Ethernet technology to Gigabit Ethernet Like-wise, the need to support videoconferencing may require a backbone upgrade from CAT 5 to fiber-optic cable

If you recall the cabling and hardware required for different networking technologies (asexplained in Chapters 3 and 6), you get an idea of how far-reaching a backbone upgrade can

be For example, to convert from Token Ring to Ethernet, you must replace or upgrade ity equipment such as hubs (or MAUs), switches, and routers In addition, you must replacethe NIC in every workstation and printer on the network and change the configuration foreach device so that it works with Ethernet rather than Token Ring For a small network, thiseffort may not be more than a weekend’s work For a network of thousands of users, such anupgrade requires the services of a dedicated team

connectiv-Because backbone upgrades are expensive and time-consuming, the first step in approachingsuch a project is to justify it Will the benefits outweigh the costs? Can the upgrade wait a year

or more? If so, you might be wise to wait and find out whether a cheaper or better technicalsolution becomes available later Don’t try to wait until the technology “settles down,” becausenetworking progress never stands still On the other hand, do wait to implement brand-newtechnology until you can find out how it has worked on other networks similar to your own oruntil the manufacturer eliminates most of the bugs

The second step is to determine which kind of backbone design to implement To make thisdecision, you must analyze the future capacity needs of your network, decide whether youwant a distributed or collapsed backbone, determine whether you want to rely on switches orrouters, decide whether to use subnetting and to what extent, and so on Although some ofthese predictions will be guesswork, you can minimize the variables by examining the history

of your organization’s growth and needs

After designing your backbone upgrade, develop a project plan to accomplish the upgrade.Given that you don’t upgrade your backbone every day, you might want to contract this work

to a firm that specializes in network design and upgrades In that case, you will draft an RFP(request for proposal) to specify what that contractor should do Regardless of whether youemploy specialists, your project plan should include a logical process for upgrading the back-bone one section at a time (if possible) Because this process causes network outages, deter-mine how best to proceed based on users’ needs Choose a time when usage is low (such asover a holiday) to perform your upgrade

Reversing Hardware Changes

As with software changes, you should provide a way to reverse the hardware upgrade and stall the old hardware if necessary If you are replacing a faulty component or device, thisrestoration, of course, is not possible If you are upgrading a component in a device, on the otherhand, you should keep the old component safe (for example, keep NICs in static-resistantcontainers) and nearby Not only might you need to put it back in the device, but you mightalso need to refer to it for information For example, if you have not documented the neces-sary jumper settings for an interface card in a switch, the old card might indicate the jumpersettings needed on your new card Even if the device seems to be operating well with the newcomponent, keep the old component for a while, especially if it is the only one of its kind atyour organization

rein-Chapter Summary

◆ Project management is the practice of managing staff, budget, timelines, and other

resources and variables so as to complete a specific goal within given bounds The personwho designs the project plan and oversees the project is the project manager A project

needs a plan, and also participants, funding, a specific means of communication, definitiveprocesses, contingency plans, and a testing and evaluation phase

◆ The first step in project management is to conduct a feasibility study A feasibilitystudy determines whether a proposed project fits within an organization’s budget,time, and staff restrictions It also attempts to weigh the benefits and costs of under-taking a project

◆ A needs assessment is the process of clarifying the reasons and objectives for a posed change It involves interviewing users and other stakeholders and comparingtheir perceptions to factual data In addition, it may involve analyzing network

pro-Chapter 15 697

CHAPTER SUMMARY

baseline data Your goal in performing a needs assessment is to decide whether thechange is worthwhile and necessary and to determine the appropriate scope andnature of the change.

◆ Project goals help keep a project on track They are also necessary later for ing whether a project was successful Project managers typically begin with a broadgoal, then narrow it down into specific goals that contribute to the larger goal

evaluat-◆ A project plan describes how the details of a managed project are organized Itdivides tasks and subtasks, dependencies, resource allocation, timelines, and mile-stones Project plans may take the form of a simple text or spreadsheet document forsmall projects Larger projects, however, often require the use of project managementsoftware (such as Microsoft Project)

◆ When implementing a major network change, communicate with users about howtheir access to the network will be affected; for how long their access to the networkwill be affected; how their data will be protected during the change; whether youwill provide any means for users to access the network during the change; andwhether the change requires them to learn new skills

◆ The best way to evaluate a large-scale network or systems implementation is to firsttest it on a small scale on a pilot network Although a pilot network differs from theenterprise-wide network, it should mimic it closely enough to represent the largernetwork’s hardware, software, connectivity, unique configurations, and load

◆ Network management involves assessing, monitoring, and maintaining networkdevices and connections

◆ Baselining includes keeping a history of network performance, the physical topology,logical topology, number of devices on the network, operating systems and protocols

in use, and number and type of applications in use In other words, it provides acomplete picture of the network’s current state Baselining provides the basis fordetermining what types of changes might improve the network and for later evaluat-ing how successful the improvements were

◆ Assessing a network’s status on an ongoing basis includes performance management,

or monitoring how well links and devices are keeping up with the demands placed

on them, and fault management, or the detection and signaling of device, link, orcomponent faults

◆ Network management applications typically use SNMP (Simple Network ment Protocol) to communicate with agents running on managed devices Agentsmay report information on a device’s components or status (such as utilization orperformance)

Manage-◆ An asset management system includes an inventory of the total number of nents on the network as well as each device’s configuration files, model number, ser-ial number, location on the network, and technical support contact In addition, itrecords every piece of software purchased by your organization, its version number,vendor, and technical support contact

compo-◆ A patch is an enhancement or improvement to a part of a software application, often tributed at no charge by software vendors to fix a bug in their code or to add slightly morefunctionality Patches differ from revisions and software upgrades because they change

dis-only part of the software application, leaving most of the code untouched

◆ Make it a policy to keep informed about patches to your network software, whetherthey involve the operating system, an application, or a client software If you work in

a large organization with several servers, routers, and other devices, you may want toassign one network administrator to manage patches for the servers, another to

manage patches for the printers, and so on

◆ A software upgrade represents a major change to the existing code, which may ormay not be offered free from a vendor and may or may not be comprehensive

enough to substitute for the original application An upgrade to the client softwarereplaces the existing client software so as to add functionality and fix bugs found inthe previous version

◆ Before upgrading client software, carefully read the instructions that accompany theupgrade to find out how best to apply it, whether it depends on any previous

upgrades, whether it requires any special preparation, and how its changes will affectusers Client upgrades typically overwrite some system files on the workstation, sotheir installation may affect other applications adversely

◆ Like client upgrades, application upgrades consist of modifications to all or part of

an application that are designed to enhance functionality or fix problems with thesoftware Application upgrades, however, affect software applications shared by

clients on the network

◆ Perhaps the most critical type of software upgrade you’ll perform comprises an

upgrade to your network operating system This effort usually involves significant,potentially drastic, changes to the operation of your servers and clients As such, itrequires plenty of forethought, product research, and rigorous testing before you

implement it In fact, for any network with more than a few users, you should createand follow a project plan for this undertaking

◆ The process of upgrading an NOS should include research, proposal, evaluation,

training, pre-implementation, implementation, and post-implementation phases

◆ You should plan for the possibility that a software upgrade might harm your existingsystem (or systems), and be prepared to reverse the process The restoration of a pre-vious version of software after an attempted upgrade is known as backleveling

◆ Hardware and physical plant changes may be required when your network has lems More often, however, they are performed as part of a move to increase capac-ity, improve performance, or add functionality to the network

prob-◆ Research, evaluate, and test any unfamiliar piece of equipment you intend to add orupgrade on your network, even if it is manufactured by a vendor that supplies much

of your other hardware The process of implementing a hardware upgrade is very

similar to that of carrying out a software upgrade, including notifying users and

preparing to bring the system down during the change

Chapter 15 699

CHAPTER SUMMARY

◆ Cabling upgrades are simpler and less error-prone if a network’s cable plant is welldocumented Also make sure to document new cable infrastructure after makingchanges When embarking on a major cabling upgrade, such as a backbone replace-ment, it is advisable to upgrade the infrastructure in phases.

◆ The most comprehensive and complex upgrade involving network hardware is abackbone upgrade The network backbone serves as the main conduit for data onLANs and WANs, connecting major routers, servers, and/or switches A backboneupgrade not only requires a great deal of time to plan, but also the efforts of severalstaff members (and possibly contractors) and a significant investment

◆ You should provide a way to reverse a hardware upgrade and replace it with the oldhardware If you are upgrading a component in a device, keep the old componentsafe (for example, keep NICs in static-resistant containers) and nearby Not onlymight you need to put it back in the device, but you might also need to refer to it forinformation

Key Terms

agent—A software routine that collects data about a managed device’s operation and provides

it to the network management application running on the console

asset management—The process of identifying and tracking an organization’s assets, such as

hardware and software

backleveling—The process of reverting to a previous version of a software application after

attempting to upgrade it

bug—A flaw in software or hardware that causes it to malfunction.

contingency planning—The process of identifying steps that minimize the risk of unforeseen

cir-cumstances endangering the quality or timeliness of the project’s goals

fault management—The detection and signaling of device, link, or component faults.

feasibility study—A study that determines the costs and benefits of a project and attempts to

pre-dict whether the project will result in a favorable outcome (for example, whether it will achieve itsgoal without imposing excessive cost or time burdens on the organization)

Gantt chart—A popular method of depicting when projects begin and end along a horizontal

time-line

Management Information Base—See MIB.

MIB (Management Information Base)—A database used in network management that contains a

device’s definitions of managed objects and their data

milestone—A reference point that marks the completion of a major task or group of tasks in a

pro-ject and contributes to measuring the propro-ject’s progress

MRTG (Multi Router Traffic Grapher)—A command-line utility that uses SNMP to poll devices,

collects data in a log file, and then generates HTML-based views of the data MRTG is freely tributed software originally written by Tobias Oetiker, a networking professional who in the early1990s saw a need for a tool to regularly measure the status of his organization’s WAN link

dis-Multi Router Traffic Grapher—See MRTG.

needs assessment—The process of clarifying the reasons and objectives for a proposed change to

determine whether the change is worthwhile and necessary, and to elucidate the scope and nature ofthe proposed change

network management—The assessment, monitoring, and maintenance of the devices and

connec-tions on a network

patch—A correction, improvement, or enhancement to part of a software application, often

distrib-uted at no charge by software vendors to fix a bug in their code or to add slightly more ity

functional-performance management—The ongoing assessment of how well network links, devices, and

com-ponents keep up with demands on them

pilot network—A small-scale network that stands in for the larger network A pilot network may

be used to evaluate the effects of network changes or additions

polling—A network management application’s regular collection of data from managed devices predecessor—A task in a project that must be completed before other tasks can begin.

project management—The practice of managing staff, budget, timelines, and other resources

and variables to complete a specific goal within given bounds

project plan—The way in which details of a managed project (for example, the timeline and

the significant tasks) are organized Some project plans are created via special project planningsoftware, such as Microsoft Project

service pack—A significant patch to one of the Microsoft Windows operating systems.

Simple Network Management Protocol—See SNMP.

SNMP (Simple Network Management Protocol)—An Application layer protocol in the

TCP/IP suite used to convey data regarding the status of managed devices on a network

sponsor—A person in a position of authority who supports a project and who can lobby for

budget increases necessary to complete the project, appeal to a group of managers to extend aproject’s deadline, and assist with negotiating vendor contracts

stakeholder—Any person who may be affected by a project, for better or for worse A

stake-holder may be a project participant, user, manager, or vendor

upgrade—A major change to the existing code in a software application, which may or may

not be offered free from a vendor, and may or may not be comprehensive enough to substitutefor the original application

Chapter 15 701

KEY TERMS

6. True or false? Migrating from a Token Ring network to Ethernet is an example of abackbone upgrade.

7. True or false? A contingency plan is a popular method for depicting when projects

begin and end along a horizontal timeline

8. True or false? You do not have to worry about downtime or notifying users when

adding a new hub or access point, because it cannot affect anyone until it is actually inuse

9. True or false? The detection and signaling of device, link, or component faults is

known as performance management

10. True or false? A small-scale network that stands in for a larger network is sometimescalled a pilot network

11. A(n) _ is a reference point that marks the completion of

a major task or group of tasks in the project and contributes to measuring the project’sprogress

12. A(n) _ outlines the costs and benefits of the project andattempts to predict whether it will result in a favorable outcome

13. _ is the process of identifying steps that minimize therisk of unforeseen events that could affect the quality or timeliness of the project’s

Examination Objectives

Appendix A

This book covers all of the Network+ examination objectives, which were released by CompTIA (the Computing Technology Industry Association)

in 2005 The official list of objectives is available at CompTIA’s Web site,

www.comptia.org For your reference, the following table lists each exam

objective and the chapter of this book that explains the objective, plus the amount of the exam that will cover each certification domain Each objec- tive belongs to one of four domains (or main topics) of networking exper- tise For example, the objective of recognizing an RJ-45 connector belongs

to the “Media and Topologies” domain, which accounts for 20% of the exam’s content.

Table A-1 Network+ Examination Objectives—Media and Topologies

1.1 Recognize the following logical or physical network topologies given a

diagram, schematic or description:

1.2 Specify the main features of 802.2 (Logical Link Control), 802.3

(Ethernet), 802.5 (token ring), 802.11 (wireless), and FDDI (Fiber Distributed

Data Interface) networking technologies, including:

Access method (CSMA / CA (Carrier Sense Multiple Access / Collision

Avoidance) and CSMA / CD (Carrier Sense Multiple Access / Collision

1.3 Specify the characteristics (For example: speed, length, topology, and cable

type) of the following cable standards:

1000BASE-T, 1000BASE-CX, 1000BASE-SX and 1000BASE-LX 3

1.4 Recognize the following media connectors and describe their uses: