Network+ 2005 In Depth (P8) pps

ARP Address Resolution Protocol—A core protocol in the TCP/IP suite that belongs in the Network layer of the OSI Model.. BOOTP Bootstrap Protocol—An Application layer protocol in the TCP

◆ If a computer runs the Windows 98, Me, 2000, 2003, or XP operating system, isconfigured to use DHCP and cannot locate a DHCP server, it can be assigned an

IP address and subnet mask through APIPA (Automatic Private IP Addressing).This configuration allows the computer to communicate with other computers onthe same subnet only

◆ A socket is a logical address assigned to a specific process running on a host Itforms a virtual circuit between the processes on two networked hosts The socket’saddress represents a combination of the host’s IP address and the port number asso-ciated with a process

◆ IPv6 (IP version 6) is the latest version of IP Its addresses are composed of eight16-bit fields and total 128 bits The larger address size results in up to 296available

IP addresses IPv6 provides several other benefits over IPv4, including a more cient header, better overall security, better prioritization allowances, and automatic

effi-IP address configuration effi-IPv6 is not yet widely implemented

◆ Every host is identified by a host name and belongs to a domain A domain is a group ofhosts that share a domain name and have part of their IP addresses in common

◆ Every domain is identified by its domain name Usually, a domain name is associatedwith a company or other type of organization, such as a university or military unit.Domain names must be reserved with an ICANN-approved domain registrar

◆ DNS (Domain Name System) is a hierarchical way of tracking domain names andtheir addresses The DNS database does not rely on one file or even one server, butrather is distributed over several key computers across the Internet to prevent cata-strophic failure if one or a few computers go down

◆ Name servers (or DNS servers) contain databases of names and their associated IPaddresses If one name server cannot resolve the IP address, the query passes to ahigher-level name server Each name server manages a group of machines called azone DNS relies on the hierarchical zones to distribute naming information

◆ When one host needs to communicate with another host, it must first find its nameserver Large organizations often maintain a primary and a secondary name server tohelp ensure Internet connectivity You need to specify a name server’s IP address inthe TCP/IP properties of a workstation so that the workstation will know whichmachine to query when looking up a name

Some key TCP/IP Application layer protocols include Telnet (for logging intohosts), FTP and TFTP (for transferring files between hosts), NTP (for synchroniz-ing time between hosts), NNTP (for storage and distribution of newsgroup mes-sages), and PING (for sending echo requests and echo replies that can indicatewhether a host is responding)

◆ IPX/SPX (Internetwork Packet Exchange/Sequenced Packet Exchange) was used byNovell for its early versions of the NetWare NOS IPX/SPX is required for interop-erability with LANs running NetWare versions 3.2 and lower IPX/SPX is a suite ofprotocols that reside at different layers of the OSI Model The IPX protocol handlesnetwork addressing information, making IPX/SPX routable

◆ IPX addresses contain two parts: the network address and the node address The

network address must be an 8-bit hexadecimal address The node address is equal to

a device’s MAC address

◆ NetBEUI is a protocol that consumes few network resources, provides error tion, and requires little configuration But it can support only 254 connections anddoes not allow for good security Furthermore, because NetBEUI lacks a Networklayer, it is not routable and therefore unsuitable for large networks

correc-◆ WINS (Windows Internet Naming Service) is a service used on Windows systems

to map IP addresses to NetBIOS names

◆ AppleTalk is the protocol suite originally used to interconnect Macintosh ers Today’s Macintosh computers can still communicate via AppleTalk, but use

comput-TCP/IP as their default protocol suite

Key Terms

Address Resolution Protocol—See ARP.

address resource record—A type of DNS data record that maps the IP address of an

Inter-net-connected device to its domain name

alias—A nickname for a node’s host name Aliases can be specified in a local host file.

anycast address—A type of address specified in IPv6 that represents a group of interfaces, any

one of which (and usually the first available of which) can accept a transmission At this time,anycast addresses are not designed to be assigned to hosts, such as servers or workstations, butrather to routers

AppleTalk—The protocol suite used to interconnect Macintosh computers Although

AppleTalk was originally designed to support peer-to-peer networking among Macintoshes, itcan now be routed between network segments and integrated with NetWare- or Microsoft-based networks

AppleTalk network number—A unique 16-bit number that identifies the network to which

an AppleTalk node is connected

AppleTalk node ID—A unique 8-bit or 16-bit number that identifies a computer on an

AppleTalk network

AppleTalk zone—A logically defined group of computers on an AppleTalk network.

ARP (Address Resolution Protocol)—A core protocol in the TCP/IP suite that belongs in

the Network layer of the OSI Model ARP obtains the MAC (physical) address of a host, ornode, and then creates a local database that maps the MAC address to the host’s IP (logical)address

ARP cache—See ARP table.

Chapter 4 183

KEY TERMS

ARP table—A database of records that map MAC addresses to IP addresses The ARP table

is stored on a computer’s hard disk where it is used by the ARP utility to supply the MACaddresses of network nodes, given their IP addresses

binding—The process of assigning one network component to work with another.

BOOTP (Bootstrap Protocol)—An Application layer protocol in the TCP/IP suite that uses

a central list of IP addresses and their associated devices’ MAC addresses to assign IP addresses

to clients dynamically BOOTP was the precursor to DHCP

Bootstrap Protocol—See BOOTP.

DHCP (Dynamic Host Configuration Protocol)—An Application layer protocol in the

TCP/IP suite that manages the dynamic distribution of IP addresses on a network UsingDHCP to assign IP addresses can nearly eliminate duplicate-addressing problems

diskless workstation—A workstation that doesn’t contain a hard disk, but instead relies on a

small amount of read-only memory to connect to a network and to pick up its system files

DNS (Domain Name System or Domain Name Service)—A hierarchical way of tracking

domain names and their addresses, devised in the mid-1980s The DNS database does not rely

on one file or even one server, but rather is distributed over several key computers across theInternet to prevent catastrophic failure if one or a few computers go down DNS is a TCP/IPservice that belongs to the Application layer of the OSI Model

domain name—The symbolic name that identifies a domain Usually, a domain name is

asso-ciated with a company or other type of organization, such as a university or military unit

Domain Name Service—See DNS.

Domain Name System—See DNS.

dotted decimal notation—The shorthand convention used to represent IP addresses and make

them more easily readable by humans In dotted decimal notation, a decimal number between

0 and 255 represents each binary octet A period, or dot, separates each decimal

dynamic address—An IP address that is assigned to a device through DHCP and may change

when the DHCP lease expires or is terminated

dynamic ARP table entry—A record in an ARP table that is created when a client makes an

ARP request that cannot be satisfied by data already in the ARP table

Dynamic Host Configuration Protocol—See DHCP

dynamic IP address—An IP address that is assigned to a device upon request and may change

over time BOOTP and DHCP are two ways of assigning dynamic IP addresses

Dynamic Ports—TCP/IP ports in the range of 49152 through 65535, which are open for use

with-out requiring administrative privileges on a host or approval from IANA

echo reply—The response signal sent by a device after another device pings it.

echo request—The request for a response generated when one device pings another device external network number—Another term for the network address portion of an IPX/SPX

address

File Transfer Protocol—See FTP.

Format Prefix—A variable-length field at the beginning of an IPv6 address that indicates what

type of address it is (for example, unicast, anycast, or multicast)

FTP (File Transfer Protocol)—An Application layer protocol used to send and receive files

via TCP/IP

hop—A term used to describe each trip a unit of data takes from one connectivity device to

another Typically, “hop” is used in the context of router-to-router communications

host file—A text file that associates TCP/IP host names with IP addresses.

host name—A symbolic name that describes a TCP/IP device.

hosts—Name of the host file used on UNIX, Linux, and Windows systems On a UNIX- or

Linux-based computer, hosts is found in the /etc directory On a Windows-based computer, it

is found in the %systemroot%\system32\drivers\etc folder

ICMP (Internet Control Message Protocol)—A core protocol in the TCP/IP suite that

noti-fies the sender that something has gone wrong in the transmission process and that packetswere not delivered

IGMP (Internet Group Management Protocol or Internet Group Multicast Protocol)—A

TCP/IP protocol used to manage multicast transmissions Routers use IGMP to determinewhich nodes use IGMP to join or leave a multicast group

Internet Control Message Protocol—See ICMP.

Internet Group Management Protocol—See IGMP.

Internet Group Multicast Protocol—See IGMP.

internetwork—To traverse more than one LAN segment and more than one type of network

through a router

Internetwork Packet Exchange—See IPX.

Internetwork Packet Exchange/Sequenced Packet Exchange—See IPX/SPX.

IP datagram—The IP portion of a TCP/IP frame that acts as an envelope for data, holding

information necessary for routers to transfer data between subnets

IP next generation—See IPv6.

IPv4LL (IP version 4 Link Local)—A protocol that manages automatic address assignment

among locally connected nodes IPv4LL is part of the Zeroconf group of protocols

Chapter 4 185

KEY TERMS

ifconfig—A TCP/IP configuration and management utility used with UNIX and Linux

sys-tems

ipconfig—The utility used to display TCP/IP addressing and domain name information in the

Windows NT, Windows 2000, and Windows XP operating systems

IPng—See IPv6.

IPv4 (IP version 4)—The current standard for IP addressing that specifies 32-bit addresses

composed of four octets

IPv6 (IP version 6)—A newer standard for IP addressing that will replace the current IPv4

(IP version 4) Most notably, IPv6 uses a newer, more efficient header in its packets and allowsfor 128-bit source and destination IP addresses The use of longer addresses will allow for manymore IP addresses to be in circulation

IPX (Internetwork Packet Exchange)—A core protocol of the IPX/SPX suite that operates

at the Network layer of the OSI Model and provides routing and internetwork services, lar to IP in the TCP/IP suite

simi-IPX address—An address assigned to a device on an simi-IPX/SPX-based network.

IPX/SPX (Internetwork Packet Exchange/Sequenced Packet Exchange)—A protocol

orig-inally developed by Xerox, then modified and adopted by Novell in the 1980s for the NetWarenetwork operating system

label—A character string that represents a domain (either top-level, second-level, or

third-level)

lease—The agreement between a DHCP server and client on how long the client can use a

DHCP-assigned IP address DHCP services can be configured to provide lease terms equal toany amount of time

loopback address—An IP address reserved for communicating from a node to itself (used

mostly for troubleshooting purposes) The loopback address is always cited as 127.0.0.1,although in fact, transmitting to any IP address whose first octet is “127” will contact the orig-inating device

loopback test—An attempt to contact one’s own machine for troubleshooting purposes In

TCP/IP-based networking, a loopback test can be performed by communicating with an IPaddress that begins with an octet of 127 Usually, this means pinging the address 127.0.0.1

multicast address—A type of address in the IPv6 that represents multiple interfaces, often on

multiple nodes An IPv6 multicast address begins with the following hexadecimal field: FF0x, where x is a character that identifies the address’s group scope.

multicasting—A means of transmission in which one device sends data to a specific group of

devices (not necessarily the entire network segment) in a point-to-multipoint fashion Multicastingcan be used for videoconferencing over the Internet, for example

multiprotocol network—A network that uses more than one protocol.

name server—A server that contains a database of TCP/IP host names and their associated

IP addresses A name server supplies a resolver with the requested information If it cannotresolve the IP address, the query passes to a higher-level name server

name space—The database of Internet IP addresses and their associated names distributed over

DNS name servers worldwide

net mask—See subnet mask.

NetBEUI (NetBIOS Enhanced User Interface)—The Microsoft adaptation of the IBM

BIOS protocol NetBEUI expands on NetBIOS by adding a Transport layer component BEUI is a fast and efficient protocol that consumes few network resources, provides excellenterror correction, and requires little configuration

Net-NetBIOS (Network Basic Input Output System)—A protocol designed by IBM to provide

Transport and Session layer services for applications running on small, homogeneous networks

NetBIOS Enhanced User Interface—See NetBEUI.

Network Basic Input Output System—See NetBIOS.

network class—A classification for TCP/IP-based networks that pertains to the network’s

potential size and is indicated by an IP address’s network ID and subnet mask Networkclasses A, B, and C are commonly used by clients on LANs; network classes D and E arereserved for special purposes

network ID—The portion of an IP address common to all nodes on the same network or

sub-net

Network News Transport Protocol—See NNTP.

Network Time Protocol—See NTP.

newsgroup—An Internet-based forum for exchanging messages on a particular topic

News-groups rely on NNTP for the collection and dissemination of messages

NNTP (Network News Transport Protocol)—An Application layer protocol in the TCP/IP

suite which facilitates the exchange of newsgroup messages, or articles, between multiple serversand users

NTP (Network Time Protocol)—A simple Application layer protocol in the TCP/IP suite

used to synchronize the clocks of computers on a network NTP depends on UDP for port layer services

Trans-octet—One of the four 8-bit bytes that are separated by periods and together make up an IP

address

Packet Internet Groper—See PING.

ping—To send an echo request signal from one node on a TCP/IP-based network to another,

using the PING utility See also PING.

Chapter 4 187

KEY TERMS

PING (Packet Internet Groper)—A TCP/IP troubleshooting utility that can verify that

TCP/IP is installed, bound to the NIC, configured correctly, and communicating with thenetwork PING uses ICMP to send echo request and echo reply messages that determine thevalidity of an IP address

port number—The address on a host where an application makes itself available to incoming

data

RARP (Reverse Address Resolution Protocol)—A core protocol in the TCP/IP suite that

belongs in the Network layer of the OSI Model RARP relies on a RARP table to associatethe IP (logical) address of a node with its MAC (physical) address RARP can be used to sup-ply IP addresses to diskless workstations

Registered Ports—TCP/IP ports in the range of 1024 to 49151 These ports are accessible to

network users and processes that do not have special administrative privileges Default ments of these ports must be registered with IANA

assign-release—The act of terminating a DHCP lease.

Rendezvous—Apple Computer’s implementation of the Zeroconf group of protocols resolver—Any host on the Internet that needs to look up domain name information resource record—The element of a DNS database stored on a name server that contains

information about TCP/IP host names and their addresses

Reverse Address Resolution Protocol—See RARP.

root server—A DNS server maintained by ICANN and IANA that is an authority on how to

contact the top-level domains, such as those ending with com, edu, net, us, and so on.ICANN oversees the operation of 13 root servers around the world

routable—Protocols that can span more than one LAN because they carry Network layer and

addressing information that can be interpreted by a router

Sequenced Packet Exchange—See SPX.

socket—A logical address assigned to a specific process running on a computer Some sockets

are reserved for operating system functions

SPX (Sequenced Packet Exchange)—One of the core protocols in the IPX/SPX suite SPX

belongs to the Transport layer of the OSI Model and works in tandem with IPX to ensure thatdata are received whole, in sequence, and error free

static ARP table entry—A record in an ARP table that someone has manually entered using

the ARP utility Static ARP table entries remain the same until someone manually modifiesthem with the ARP utility

static IP address—An IP address that is manually assigned to a device and remains constant

until it is manually changed

subnet—A part of a network in which all nodes shares a network addressing component and

a fixed amount of bandwidth

subnet mask—A 32-bit number that, when combined with a device’s IP address, indicates what

kind of subnet the device belongs to

subnetting—The process of subdividing a single class of network into multiple, smaller

net-works

subprotocols—Small, specialized protocols that work together and belong to a protocol suite switch—The letters or words added to a command that allow you to customize a utility’s out-

put Switches are usually preceded by a hyphen or forward slash character

TCP (Transmission Control Protocol)—A core protocol of the TCP/IP suite TCP belongs

to the Transport layer and provides reliable data delivery services

TCP/IP (Transmission Control Protocol/Internet Protocol)—A suite of networking

pro-tocols that includes TCP, IP, UDP, and many others TCP/IP provides the foundation for data exchange across the Internet

TCP/IP core protocols—The major subprotocols of the TCP/IP suite, including IP, TCP, and

UDP

Telnet—A terminal emulation protocol used to log on to remote hosts using the TCP/IP

pro-tocol Telnet resides in the Application layer of the OSI Model

TFTP (Trivial File Transfer Protocol)—A TCP/IP Application layer protocol that enables

file transfers between computers Unlike FTP, TFTP relies on UDP at the Transport layer anddoes not require a user to log on to the remote host

Time to Live—See TTL.

TLD (top-level domain)—The highest-level category used to distinguish domain names—

for example, org, com, net A TLD is also known as the domain suffix

top-level domain—See TLD.

Transmission Control Protocol—See TCP.

Transmission Control Protocol/Internet Protocol—See TCP/IP.

Trivial File Transfer Protocol—See TFTP.

TTL (Time to Live)—A number that indicates the maximum time that a datagram or packet

can remain on the network before it is discarded Although this field was originally meant torepresent units of time, on modern networks it represents the number of router hops a data-gram has endured The TTL for datagrams is variable and configurable, but is usually set at 32

or 64 Each time a datagram passes through a router, its TTL is reduced by 1 When a routerreceives a datagram with a TTL equal to 1, the router discards that datagram

Chapter 4 189

KEY TERMS

UDP (User Datagram Protocol)—A core protocol in the TCP/IP suite that sits in the

Trans-port layer of the OSI Model UDP is a connectionless transTrans-port service

unicast address—A type of IPv6 address that represents a single interface on a device An IPv6

unicast address begins with either FFC0 or FF80

User Datagram Protocol—See UDP.

Well Known Ports—TCP/IP port numbers 0 to 1023, so named because they were long ago

assigned by Internet authorities to popular services (for example, FTP and Telnet), and aretherefore well known and frequently used

Windows Internet Naming Service—See WINS.

WINS (Windows Internet Naming Service)—A service that resolves NetBIOS names with

IP addresses WINS is used exclusively with systems that use NetBIOS—therefore, it is found

on Windows-based systems

Zeroconf (Zero Configuration)—A collection of protocols designed by the IETF to simplify

the setup of nodes on a TCP/IP network Zeroconf assigns a node an IP address, resolves thenode’s host name and IP address without requiring a DNS server, and discovers services, such

as print services, available to the node, also without requiring a DNS server

a. Network Time Protocol

b. File Transfer Protocol

c. Address Resolution Protocol

d. Internet Control Message Protocol

3. _ contain databases of associated names and IP

addresses and provide this information to resolvers on request

a. Dynamic Host Configuration Protocol

b. Windows Internet Naming Service

c. Network News Transport Protocol

d. Internet Packet Exchange Protocol

5. _ is the process of assigning one network component towork with another

a. Subnetting

b. Multicasting

c. Binding

d. IP addressing

6. True or false? All protocols are routable

7. True or false? TCP ensures reliable delivery through sequencing and checksums

8. True or false? TCP is a connectionless transport device

9. True or false? Every process on a machine is assigned a port number

10. True or false? IPv6 addresses are composed of eight 16-bit fields and total 32 bits

11. _ allows one device to send data to a specific group ofdevices

12. A(n) _ is a special 32-bit number that, when combinedwith a device’s IP address, informs the rest of the network about the segment or net-work to which it is attached

Chapter 4 191

REVIEW QUESTIONS

13. _ are any hosts on the Internet that need to look updomain name information.

14. _ is a terminal emulation protocol used to log on toremote hosts using the TCP/IP protocol suite

15. The _ is a simple Application layer protocol used tosynchronize the clocks of computers on a network

Networking Hardware

Chapter 5

After reading this chapter and completing the exercises, you will be able to:

■ Identify the functions of LAN connectivity hardware

■ Install and configure a network interface card (NIC, or network adapter)

■ Identify problems associated with connectivity hardware

■ Describe the factors involved in choosing a NIC, hub, switch, or router

■ Discuss the functions of repeaters, hubs, bridges, switches, routers, and gateways, and the OSI Model layers at which they operate

■ Describe the uses and types of routing protocols

mail: Mail trucks, airplanes, and delivery staff serve as the transmission system that movesinformation from place to place Machines and personnel at the post office interpret addresses

on the envelopes and either deliver the mail to a transfer point or to your home Inefficiencies

in mail delivery, such as letters being misdirected to the wrong transfer point, frustrate boththe sender and the receiver of the mail and increase the overall cost of delivery

In data networks, the task of directing information efficiently to the correct destination is dled by hubs, routers, bridges, and switches In this chapter, you will learn about these devicesand their roles in managing data traffic Material in this chapter relates mostly to functionsoccurring in the Data Link and Network layers of the OSI Model Some material also relates

han-to the Physical layer You will learn the concepts involved in moving data from place han-to place,including issues related to switching and routing protocols You will also see pictures of thehardware—hubs, switches, bridges, and routers—that make data transfer possible (It’s impor-tant for you to have an accurate mental image of this equipment because, in a cluttered datacloset, it may prove difficult to identify the hardware underneath the wiring.) In addition, youwill learn all about network interface cards, which serve as the workstation’s link to the net-work and are often the source of connectivity problems

NICs (Network Interface Cards)

Network interface cards (also called NICs, network adapters, or network cards) are ity devices that enable a workstation, server, printer, or other node to receive and transmit dataover the network media Nearly all NICs contain a data transceiver, the device that transmitsand receives data signals NICs belong to both the Physical layer and Data Link layer of theOSI Model, because they apply data signals to the wire and assemble or disassemble dataframes They also interpret physical addressing information to ensure data is delivered to itsproper destination In addition, they perform the routines that determine which node has theright to transmit data over a network at any given instant

connectiv-Advances in NIC technology are making this hardware smarter than ever Many can also form prioritization, network management, buffering, and traffic-filtering functions On mostnetworks, NICs do not, however, analyze information added by the protocols in Layers 3through 7 of the OSI Model For example, they could not determine whether the frames theytransmit and receive use IP or IPX datagrams Nor could they determine whether the Presen-tation layer has encrypted the data in those frames

As you learn about installing, configuring, and troubleshooting NICs, you should concentratefirst on generalities, then move on to special situations Because NICs are common to everynetworking device and every network, knowing as much as possible about them may prove to

be the most useful tool you have at your disposal

Types of NICs

Before you order or install a NIC in a network device, you need to know what type of face the device uses NICs come in a variety of types depending on:

inter-◆ The access method (for example, Ethernet versus Token Ring)

◆ Network transmission speed (for example, 100 Mbps versus 1 Gbps)

◆ Connector interfaces (for example, RJ-45 versus SC)

◆ Type of compatible motherboard or device (for example, PCI)

◆ Manufacturer (popular NIC manufacturers include 3Com, Adaptec, D-Link, IBM,Intel, Kingston, Linksys, Netgear, SMC, and Western Digital, to name just a few)The following section describes one category of NICs, those that are installed on an expansionboard inside a computer

Internal Bus Standards

If you have worked with PCs or studied for CompTIA’s A+ exam, you are probably familiar

with the concept of a bus A computer’s bus is the circuit, or signaling pathway, used by the

motherboard to transmit data to the computer’s components, including its memory, processor,

hard disk, and NIC (A computer’s bus may also be called its system bus or main bus.) Buses

differ according to their capacity The capacity of a bus is defined principally by the width ofits data path (expressed in bits) and its clock speed (expressed in MHz) A data path sizeequals the number of bits that it can transmit in parallel at any given time In the earliest PCs,buses had an 8-bit data path Later, manufacturers expanded buses to handle 16 bits of data,then 32 bits Most new desktop computers use buses capable of exchanging 64 bits of data,and some are even capable of 128 bits As the number of bits of data that a bus can handleincreases, so too does the speed of the devices attached to the bus

A computer’s bus can be expanded to include devices other than those found on the

mother-board The motherboard contains expansion slots, or openings with multiple electrical

con-tacts, that allow devices such as NICs, modems, or sound cards to connect to the computer’s

expanded bus The devices are found on a circuit board called an expansion card or expansion board Inserting an expansion board into an expansion slot establishes an electrical connection

between the expansion board and the motherboard Thus, the device connected to the sion board becomes connected to the computer’s main circuit and part of its bus With expan-sion boards connected to its main circuit, a computer can centrally control the device

expan-Chapter 5 195

NICS (NETWORK INTERFACE CARDS)

NET+

1.6

Multiple bus types exist, and to become part of a computer’s bus, an expansion board must usethe same bus type By far the most popular expansion board NIC is one that uses a PCI bus.

PCI (Peripheral Component Interconnect) is a 32- or 64-bit bus with a 33- or 66-MHz clock

speed whose maximum data transfer rate is 264 MBps Intel introduced the first version of PCI

in 1992 The latest version, 3.0, was released in 2004 and has become the expansion card typeused for nearly all NICs in new PCs It’s characterized by a shorter connector length and a

much faster data transmission capability than previous bus types such as ISA (Industry dard Architecture), the original PC bus type, developed in the early 1980s to support an 8-bit

Stan-and later 16-bit data path Stan-and a 4.77-MHz clock speed Another advantage to PCI adapters

is that they work within both PCs and Macintosh computers, allowing an organization to dardize on one type of NIC for use with all of its workstations Figure 5-1 depicts a typicalPCI NIC

stan-A newer version of the PCI standard is PCI Express, which specifies a 64-bit bus with a

133-MHz clock speed capable of transferring data at up to 500 MBps per data path, or lane, infull-duplex transmission PCI Express, which was introduced in 2002, follows a new type ofbus design and offers several advantages over the old PCI: more efficient data transfer, supportfor quality of service distinctions, error reporting and handling, and compatibility with thecurrent PCI software Also, PCI Express cards are designed to fit into PCs that currently haveolder PCI slots (This requires the addition of a small slot behind each of two existing PCIslots The PCI Express card is then inserted into both PCI slots.) PCI Express slots varydepending on the number of lanes they support: An x1 slot supports a single lane, an x2 slotsupports two lanes, and so on Each lane offers a full-duplex throughput of 500 Mbps A PCIExpress slot can support up to 16 lanes, and an x16 slot can provide 8 Gbps throughput Com-puters such as servers that must perform fast data transfer are already using the PCI Expressstandard, and manufacturers predict that PCI Express will replace PCI in most PCs in com-

ing years PCI Express is sometimes referred to as PCIe or PCIx Figure 5-2 depicts a PCI

Express x1 NIC

You can easily determine the type of bus your PC uses by reading the documentation thatcame with the computer Someday, however, you may need to replace a NIC on a PC whosedocumentation is missing To verify the type of bus a PC uses, you can look inside the PCcase (Later in this chapter, you will learn how to open a computer case, check the computer’sbus, and install a NIC safely.) Most PCs have at least two different types of bus connections

on the same motherboard Figure 5-3 illustrates a motherboard with ISA, PCI, and PCIExpress expansion slots

If a motherboard supports more than one kind of expansion slot, refer to the NIC and PC ufacturers’ guidelines (either in print or on the Web) for information on the preferred type ofNIC If possible, you should choose a NIC that matches the most modern bus on the moth-erboard For example, if a PC supports both ISA and PCI, attempt to use a PCI NIC.Although you may be able to use the older bus and NIC types without any adverse effects, someNICs will not work in an older bus if a faster, newer bus is available on the motherboard

man-NET+

1.6