2B + D An alternative name for the Basic Rate Interface (BRI) service provided by the Integrated Services Digital Network (ISDN).
5-4-3 rule An Ethernet cabling guideline stating that an Ethernet local area network (LAN) can consist of up to five cable segments, connected by four repeaters. Up to three of those cable segments can be mixing segments.
1Base5 An early Ethernet twisted pair specification that was designed to coexist with telephone signals on the other two cable pairs. 1Base5 never achieved wide acceptance, due in part to its slow 1 Mbps transmission speed.
10Base2 Another name for the Ethernet physical layer specification (also known as thin Ethernet, ThinNet, or Cheapernet) that uses RG-58 coaxial cable in a bus topology. The “10” refers to the network’s speed of 10 Mbps, the “base” refers to the network’s baseband transmissions, and the “2” refers to the network’s maximum segment length of approximately 200 meters (actually 185 meters).
10Base5 Another name for the Ethernet physical layer specification that is also known as thick Ethernet or ThickNet, which uses RG-8 coaxial cable in a bus topology. The “10” refers to the network’s speed of 10 Mbps, the “base” refers to the network’s baseband transmissions, and the “5” refers to the network’s maximum segment length of 500 meters.
10Base-F A collective term for the three 10-Mbps Ethernet physical layer specifications that use fiber optic cable, as defined in Institute of Electrical and Electronics Engineers (IEEE) 802.3, including 10Base-FB, 10Base-FL, and 10Base-FP. The use of fiber optic cable for Ethernet networks was relatively rare until the advent of Fast Ethernet because the 10-Mbps speed limitation of the 10Base-F networks made them impractical.
10Base-FB Another name for one of three 10-Mbps Ethernet physical layer standards defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 document that use 62.5/125 multimode fiber optic cable in a star topology. 10Base-FB has a maximum segment length of 2000 meters and was intended for use as a backbone solution to connect hubs over long distances using synchronous signaling. Like the other 10Base-F specifications, it was rarely used.
10Base-FL Another name for one of three 10-Mbps Ethernet physical layer standards defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 document that use 62.5/125 multimode fiber optic cable in a star topology. 10Base-FL has a maximum segment length of 2000 meters and can connect two repeaters, two computers, or a computer to a repeater. Like the other 10Base-F specifications, it was rarely used.
10Base-FP Another name for one of three 10-Mbps Ethernet physical layer standards defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 document that use 62.5/125 multimode fiber optic cable in a star topology. 10Base-FP has a maximum segment length of 500 meters and uses a passive star coupler to connect up to 33 computers. It was designed to be the desktop fiber optic solution of the 10Base-F specifications, but, like the other 10Base-F specifications, it was rarely used.
10Base-T Another name for an Ethernet physical layer specification that uses unshielded twisted pair (UTP) cables in a star topology. The “10” refers to the network’s speed of 10 Mbps, the “base” refers to the network’s baseband transmissions, and the “T” refers to the use of twisted pair cable. The maximum cable segment length for a 10Base-T network is 100 meters.
100Base-FX Another name for a 100-Mbps Fast Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3u document that uses 62.5/125 multimode fiber optic cable in a star topology with a maximum segment length of 412 meters and runs at 100 Mbps.
100Base-T A collective term for the three 100-Mbps Ethernet physical layer specifications defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3u document and commonly known as Fast Ethernet. The three physical layer options for Fast Ethernet are 100Base-TX, 100Base-T4, and 100Base-FX.
100Base-T2 An unimplemented Fast Ethernet physical layer specification that called for the use of two of the wire pairs in a Category 3 unshielded twisted pair (UTP) cable.
100Base-T4 Another name for a 100-Mbps Fast Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3u document that uses Category 3 (CAT3) unshielded twisted pair (UTP) cable in a star topology, with a maximum segment length of 100 meters. 100Base-T4 can achieve its high speed using a lesser grade of cable because it uses all four pairs of wires in the cable, whereas other Ethernet UTP specifications, such as 100Base-TX and 10Base-T, use only two pairs. Because nearly all of the UTP cable installed today is at least Category 5 (CAT5), 100Base-T4 is seldom used, but it remains a viable alternative for sites with older cable installations.
100Base-TX Another name for a 100-Mbps Fast Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3u document that uses Category 5 (CAT5) or better unshielded twisted pair (UTP) cable in a star topology, with a maximum segment length of 100 meters. 100Base-TX achieves its high speed using only two pairs of the wires in the cable because the specification insists on the use of high-quality cable. 100Base-TX is the most popular of the Fast Ethernet specifications.
100Base-X A collective term for the 100Base-FX and 100Base-TX Fast Ethernet specifications, both of which use the 4B/5B signal encoding method.
1000Base-CX Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over 150-ohm shielded copper cable with a maximum segment length of 25 meters.
1000Base-FX Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over 62.5/125 multimode fiber optic cable with a maximum segment length of 412 meters.
1000Base-LH Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over 9/125 singlemode fiber optic cable with a maximum segment length of 10,000 meters.
1000Base-LX Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over either 9/125 singlemode fiber optic cable, with a maximum segment length of 5000 meters, or 50/125 or 62.5/125 multimode fiber optic cable with a maximum segment length of 550 meters.
1000Base-SX Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over 50/125 multimode fiber optic cable with a maximum segment length of 550 meters or 62.5/125 multimode fiber optic cable with a maximum segment length of 275 meters.
1000Base-T Another name for a 1000-Mbps Gigabit Ethernet network defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3ab document, which uses Category 5 (CAT5) or Category 5e (CAT5e) unshielded twisted pair (UTP) cable in a star topology, with a maximum segment length of 100 meters.
1000Base-X A collective term referring to all of the physical layer specifications for Gigabit Ethernet that use the 8B/10B signal encoding scheme, originally used on Fibre Channel networks. 1000Base-X includes all of the Gigabit Ethernet specifications except for 1000Base-T.
1000Base-ZX Another name for a 1000-Mbps Gigabit Ethernet physical layer specification defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z document, which runs over 9/125 singlemode fiber optic cable with a maximum segment length of 100,000 meters.
10Broad36 A seldom-used Ethernet physical layer specification that calls for broadband transmissions over 75-ohm coaxial cable.
abstract syntax The native format used by a computer to encode information generated by an application or process. The presentation layer of the Open Systems Interconnection (OSI) reference model receives data from the application in the system’s abstract syntax and is responsible for converting it to a common transfer syntax understood by both communicating systems. See also transfer syntax.
Active Directory directory service The enterprise directory service included with the Microsoft Windows Server 2003 and Microsoft Windows 2000 Server operating systems. Active Directory is a hierarchical directory service that consists of objects that represent users, computers, groups, and other network resources. The objects are arranged in a tree display that consists of hierarchical layers ranging upward from organizational units, to domains, to trees, and to forests. Objects are composed of attributes that contain information about the resource the object represents. When users log on to the network, their user names and passwords are authenticated against the Active Directory database by a computer that has been designated as a domain controller. This one single logon can grant users access to resources anywhere on the network. See also directory service.
Address Resolution Protocol (ARP) A Transmission Control Procotol/Internet Protocol (TCP/IP) protocol used to resolve the Internet Protocol (IP) addresses of computers on a local area network (LAN) into the hardware (or MAC) addresses needed to transmit data-link layer frames to them. Before transmitting an IP datagram, TCP/IP clients broadcast an ARP request message containing the IP address of the destination computer to the local network. The computer using that IP address must then respond with an ARP reply message containing its hardware address. With the information in the reply message, the computer can encapsulate the IP datagram in the appropriate data-link layer frame and transmit it to the destination system.
ad hoc A type of communication used on wireless LANs (WLANs) in which devices equipped with wireless network interface adapters communicate with each other at will. See also infrastructure topology.
ADSL See Asymmetrical Digital Subscriber Line (ADSL).
ADSL Termination Unit-Remote (ATU-R) The hardware device located at the client side of an ADSL connection. Also called a DSL transceiver or (incorrectly) a DSL modem. The ATU-R connects to the computer using either a universal serial bus (USB) port or a standard Ethernet network interface adapter. See also Asymmetrical Digital Subscriber Line (ADSL), Digital Subscriber Line Access Multiplexer (DSLAM).
American National Standards Institute (ANSI) A private, non-profit organization that administers and coordinates the U.S. voluntary standardization and conformity assessment system. ANSI is the official U.S. representative to the International Organization for Standardization (ISO), as well as several other international bodies.
ANSI See American National Standards Institute (ANSI).
AppleTalk A proprietary suite of networking protocols developed by Apple for use by its Macintosh computers. AppleTalk includes AppleShare, a file and printer-sharing solution that enables a Macintosh computer to function as a network server. AppleTalk is rarely used today because Macintosh computers now communicate using the industry-standard Transmission Control Protocol/Internet Protocol (TCP/IP) protocols.
application layer The top layer of the Open Systems Interconnection (OSI) reference model. The application layer provides the entrance point used by applications to access the networking protocol stack. Some of the protocols operating at the application layer include the Hypertext Transfer Protocol (HTTP), the Simple Mail Transfer Protocol (SMTP), the Dynamic Host Configuration Protocol (DHCP), the File Transfer Protocol (FTP), and the Simple Network Management Protocol (SNMP).
archive bit A one-bit flag included with all file systems that backup software programs use to determine whether a file has been modified. When a file is backed up, the backup software program typically resets (or strips away) its archive bit. The next time the file is modified, the archive bit is activated. The backup software can then run a job that backs up only the files with active archive bits, which reduces the time and media needed to perform the backup.
ARP See Address Resolution Protocol (ARP).
Arp.exe A command-line utility provided by the Microsoft TCP/IP client that is included with the Windows operating systems. Arp.exe enables you to display and manipulate the information stored in the cache created by the Address Resolution Protocol (ARP). By preloading the ARP cache, you can save time and network traffic by eliminating the ARP transaction that the TCP/IP client uses to resolve the IP address of each system it transmits to into a hardware address. See also Address Resolution Protocol (ARP).
Asymmetrical Digital Subscriber Line (ADSL) A point-to-point, digital wide area network (WAN) technology that uses standard telephone lines to provide consumers with high-speed Internet access, remote local area network (LAN) access, and other services. The term “asymmetric” refers to the fact that the service provides a higher transmission rate for downstream than for upstream traffic. Downstream transmission rates can be up to 8.448 Mbps, whereas upstream rates range up to 640 Kbps. See also Digital Subscriber Line (DSL).
Asynchronous Transfer Mode (ATM) A network communications technology based on 53-byte cells, designed to carry voice, data, and video traffic over local area networks (LANs) and wide area networks (WANs) at speeds ranging from 25.6 Mbps to 2.46 Gbps. ATM See Asynchronous Transfer Mode (ATM).
attachment unit interface (AUI) An interface that provides the connection between a computer and the RG-8 coaxial cable used by thick Ethernet networks. A thick Ethernet network interface adapter has a 15-pin AUI port, which is used to connect an AUI cable that runs to the RG-8 cable. The other end of the AUI cable is connected to a device called a vampire tap, which clamps onto the RG-8 cable and has teeth that pierce its protective insulation to make an electrical connection with the conductor inside. The term “attachment unit interface” is used by the Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard; the DIX Ethernet standards refer to the same components as the transceiver port and the transceiver cable.
attenuation The progressive weakening of a signal as it travels over a cable or other medium. The longer the distance a signal travels, the weaker the signal gets, until it becomes unreadable by the receiving system. On a data network, attenuation is one of the prime factors limiting the length of network cable segments. Different types of cables have different rates of attenuation. As a rule, copper cables are more prone to attenuation than fiber optic cables, and thinner copper cables are more prone to attenuation than thicker ones.
ATU-R See ADSL Termination Unit-Remote (ATU-R).
AUI See Attachment Unit Interface (AUI).
authoritative server A Domain Name System (DNS) server that has been designated as the definitive source of information about the computers in a particular domain. When resolving a computer’s DNS name into its Internet Protocol (IP) address, DNS servers consult the authoritative server for the domain in which that computer is located. Whatever information the authoritative server provides about that domain is understood by all DNS servers to be correct. See also Domain Name System (DNS).
autochanger A hardware device consisting of one or more backup drives, a media array, and a robotic mechanism that inserts media into and removes it from the drives. An autochanger is used to perform automated backups of large amounts of data.
automatic allocation An operational mode of Dynamic Host Configuration Protocol (DHCP) servers in which the server permanently assigns an Internet Protocol (IP) address and other Transmission Control Protocol/Internet Protocol (TCP/IP) configuration settings to a client from a pool of addresses. See also dynamic allocation, which assigns addresses in the same way, but reclaims them when a lease of a given duration expires, and manual allocation, which permanently assigns specific addresses to clients. See also Dynamic Host Configuration Protocol (DHCP).
backbone network A network used to connect a series of other networks together, forming an internetwork. Typically, a backbone is a high-speed local area network (LAN) used to route traffic from one horizontal LAN to another. Client workstations are typically not connected to the backbone, although servers sometimes are.
baseband network A network that uses a medium that can carry only one signal at a particular time. Most LANs are baseband networks; your local cable television (CATV) system is an example of a broadband network. See also broadband network.
Basic Rate Interface (BRI) An Integrated Services Digital Network (ISDN) service that consists of two 64-Kbps B channels plus one 16-Kbps D channel, enabling users to combine the B channels for a single 128-Kbps data pipe, or use them separately. Also called 2B+D, BRI is the primary consumer ISDN service used for Internet access and remote networking. See also B channel, D channel, Primary Rate Interface (PRI), Integrated Services Digital Network (ISDN).
B channel A 64-Kbps digital communications channel that is one of the fundamental units of service provided by the Integrated Services Digital Network (ISDN). B channels carry the actual data generated by the user’s applications. The Basic Rate Interface (BRI) ISDN service consists of two B channels plus one 16-Kbps D channel; the Primary Rate Interface (PRI) service consists of 23 B channels and one 64-Kbps D channel. See also Integrated Services Digital Network (ISDN).
bindery The server-based, flat file directory service used in Novell NetWare versions 3.2 and earlier. The bindery is a simple directory of user and group accounts used by NetWare to authenticate user access to server resources. Unlike more advanced directory services, which provide services for the entire enterprise, the NetWare bindery is specific to a single server. If a network has multiple NetWare servers, each has its own separate bindery, and users must have bindery accounts on each server they want to access.
bmp A file format commonly used to store graphic images in bitmap form.
BNC Short for bayonet-Neill- Concelman, a type of cable connector used on thin Ethernet networks.
BOOTP See Bootstrap Protocol (BOOTP). Bootstrap Protocol (BOOTP) A server application that can supply client computers with Internet Protocol (IP) addresses, other Transmission Control Protocol/Internet Protocol (TCP/IP) configuration parameters, and executable boot files. As the progenitor to the Dynamic Host Configuration Protocol (DHCP), BOOTP provides the same basic functions, except that it does not allocate IP addresses from a pool and reclaim them after a specified length of time. Administrators must supply the IP address and other settings for each computer to be configured by the BOOTP server. See also Dynamic Host Configuration Protocol (DHCP), Reverse Address Resolution Protocol (RARP).
branching tree See hierarchical star.
BRI See Basic Rate Interface (BRI).
bridge A network connectivity device that operates at the data-link layer of the Open Systems Interconnection (OSI) reference model and filters network traffic based on packets’ destination addresses. When you connect two network segments with a bridge, packets generated by the computers on one segment are only propagated to the other segment if they are addressed to a computer on that segment. The bridge learns which computers are connected to each segment by reading the source addresses in the packets it processes and storing the information in a table; this learning process is called transparent bridging. Other types of bridges can connect networks running different media or data-link layer protocols or connect two network segments at different locations using a wide area network (WAN) link.
broadband network A network that uses a medium that can carry multiple signals simultaneously, using a technique called multiplexing. The most common example of broadband communications is the typical cable television (CATV) network, which transmits the signals corresponding to dozens of TV channels over one cable. See also baseband network.
broadcast A message transmitted to all of the other computers on the local network. Data-link layer protocols have special addresses designated as broadcast addresses, which means that every computer that receives the message will read it into memory and process it. Local area networks (LANs) use broadcasts for a variety of tasks, such as to discover information about other computers on the network.
broadcast domain A collection of computers that will all receive a broadcast message transmitted by any one of the other computers. All of the computers on a local area network (LAN), for example, are in the same broadcast domain, as are the computers on two network segments connected by a bridge, because bridges always propagate broadcast transmissions. Two networks connected by a router, however, are in different broadcast domains, because routers do not propagate broadcasts. See also collision domain.
bus A network cabling topology in which each device is connected to the next device, forming a daisy chain with two ends, each of which must be terminated. See also topology.
cable television (CATV) network A private metropolitan area network (MAN) constructed and owned by a cable television company for the purpose of delivering TV signals to customers in a given region. Because the network technology they use is compatible with data networking, many CATV companies are now also in the business of providing Internet access to consumers using the same network that delivers the television service. The downstream transmission rates for a CATV Internet connection far exceed those of standard dial-ups and most other consumer Internet solutions, and the cost is usually very competitive.
Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) A variation on the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Media Access Control (MAC) method, which substitutes a system of verifications and acknowledgments for the collision detection mechanism. See also Carrier Sense Multiple Access with Collision Detection (CSMA/CD).
Carrier Sense Multiple Access with Collision Detection (CSMA/CD) The Media Access Control (MAC) mechanism used by Ethernet networks to regulate access to the network. Before they can transmit data, CSMA/CD systems listen to the network to determine if it is in use. If the network is free, the system transmits its data. However, sometimes another computer transmits at precisely the same time, causing a signal quality error or collision. Collisions are normal occurrences on Ethernet networks, and network interface adapters can detect them and compensate for them by discarding the collided packets and retransmitting them in a controlled manner.
CAT3 The Category 3 grade of unshielded twisted pair (UTP) cable that was at one time the most common medium used for telephone and data networks. New installations now use Category 5 (CAT5) cable, because it supports higher transmission speeds, although there are still some protocols that are designed specifically for use on older CAT3 networks, such as 100Base-T4 and 100VG-AnyLAN.
CAT5 The Category 5 grade of unshielded twisted pair (UTP) cable that is the current industry standard for telephone and data networking.
CAT5e Also called Category 5e or Enhanced Category 5, a relatively new grade of unshielded twisted pair (UTP) cable designed for use on data networks running at very high speeds, such as Gigabit Ethernet.
CAT6 Ratified in 2002, a grade of unshielded twisted pair (UTP) cable that provides higher performance levels than Category 5e (CAT5e) and more stringent specifications for resistance to crosstalk and system noise. CAT6 cables are rated at a frequency of 250 MHz. CAT6 is also called Category 6.
Category n A term used to specify a grade of unshielded twisted pair (UTP) cable, using standards developed by the Telecommunications Industry Association/Electronic Industries Alliance (TIA/EIA).
CATV See cable television (CATV) network.
CCITT See Comité Consultatif International Téléphonique et Télégraphique (CCITT).
CD-R A write-once/read-many (WORM) storage medium that can hold approximately 670 MB of data on a compact disc.
CD-ROM A read-only storage medium that can hold approximately 670 MB of data on a compact disc.
CD-RW A rewritable storage medium that can hold approximately 670 MB of data on a compact disc.
cell switching A type of network communications similar to packet switching. In cell switching, messages are broken up into discrete units of uniform size and transmitted to the destination.
channel service unit/data service unit (CSU/DSU) A hardware device that terminates the end of a leased line connection and provides testing and diagnostic capabilities. See also leased line.
cheapernet A slang term for a thin Ethernet (10Base2) network. At the time of its greatest popularity, cheapernet was significantly less expensive than its primary competitor, thick Ethernet (10Base5).
circuit switching A type of network communications in which two communicating systems establish a connection that remains open throughout the life of the transaction. The telephone network is an example of a circuit-switched network. After placing a call, the telephone system establishes a path through the network connecting the two telephones, and all communications follow that path until it is broken by one of the callers disconnecting. See also packet switching.
client A program designed to communicate with a server program on another computer, usually to request and receive information. The client provides the interface with which the user can view and manipulate the server data. A client can be a module in an operating system, such as the Client for Microsoft Networks in Microsoft Windows, which enables the user to access resources on the network’s other computers, or a separate application, such as a Web browser or e-mail reader.
client/server networking A computing model in which data processing tasks are distributed between clients, which request, display, and manipulate information, and servers, which supply and store information. Since each individual client is responsible for displaying and manipulating its own data, the server is relieved of a large part of the processing burden. The alternative is a mainframe or minicomputer system in which one computer performs all of the processing for all of the users, who work with terminals that do not have processors (dumb terminals).
cluster A group of two or more server computers connected together so that they function as a single unified resource, for purposes of fault tolerance, load balancing, and parallel processing. Clustering enables the server array to survive the failure of one or more computers and makes it possible to upgrade the system simply by adding additional computers to the cluster.
coaxial cable A type of cable used in various types of networking. A coaxial cable consists of two conductors, one wrapped around the other and separated by an insulating layer, enclosed in a protective sheath. The data signals are transmitted over the inner conductor that forms the solid core of the cable. The outer conductor, made of a wire mesh, functions as a ground. The two types of coaxial cable used in local area networking are called RG-8 and RG-58, also known as thick Ethernet and thin Ethernet, respectively.
collision In local area networking, a condition in which two computers transmit data at precisely the same time, and their signals both occupy the same cable, causing data loss. On some types of networks, such as Ethernet, collisions are a normal occurrence, whereas on Token Ring networks, they indicate a serious problem. Also called a signal quality error.
collision domain A group of computers in which any two that transmit at exactly the same time will cause a collision. All of the computers on a local area network (LAN) are in the same collision domain, for example, whereas the computers on two network segments connected by a bridge or a router are in two different collision domains. This is because the processing performed by routers and bridges introduces a slight delay between the generation of a packet on one segment and the propagation of the packet to the other segment.
Comité Consultatif International Téléphonique et Télégraphique (CCITT) An organization (in English, the International Telegraph and Telephone Consultative Committee) that, until 1992, developed and published international communications standards, such as those that govern modem signaling, compression, and error correction protocols. The organization is now known as the Telecommunications Standardization Sector of the International Telecommunications Union (ITU-T). The CCITT also published the document that defined the Open Systems Interconnection (OSI) reference model, called “The Basic Reference Model for Open Systems Interconnection.”
compression ratio The degree to which data can be compressed for storage on another medium, such as a backup medium. Compression ratios can range from 1:1 (no compression possible) to 8:1 or higher, depending on the format of the data stored in the individual files.
connectionless protocol A type of protocol that transmits messages to a destination without first establishing a connection with the destination system. Connectionless protocols have very little overhead and are used primarily for transactions that consist of a single request and reply. The Internet Protocol (IP) and the User Datagram Protocol (UDP) are both connectionless protocols.
connection-oriented protocol A type of protocol that transmits a series of messages to a destination to establish a connection before sending any application data. Establishing the connection ensures that the destination system is active and ready to receive data. Connection-oriented protocols are typically used to send large amounts of data, such as entire files, which must be split into multiple packets and which are useless unless every packet arrives at the destination without error. The Transmission Control Protocol (TCP) is a connection-oriented protocol.
convergence The process by which dynamic routers update their routing tables to reflect the current state of the internetwork. The primary advantage of dynamic routing is that it enables routers to modify their routing information automatically as the configuration of the network changes. For example, if a router malfunctions, the other nearby routers, after failing to receive regular updates from it, will eventually remove it from their routing tables, thus preventing computers on the network from using that router. The elapsed time between the failure of the router and its removal from the routing tables of the other routers is the convergence period.
counters The individual system attributes or processes monitored by the Performance console in Microsoft Windows 2000 and the Performance Monitor application in Microsoft Windows NT.
CRC See cyclical redundancy check.
crossover cable An unshielded twisted pair (UTP) cable in which the transmit contacts in each connector are wired to the receive contacts in the other connector. Using a crossover cable on a UTP Ethernet network eliminates the need for a hub. Crossover cables are used on small two-node networks and as a troubleshooting tool on larger networks.
crossover connection A twisted-pair network connection in which the transmit contacts at each end of a cable are wired to the receive contacts at the other end of that cable, without the use of a hub. Normally, a hub is required for a twisted-pair network, because the hub crosses the transmit and receive signals, enabling computers to communicate with each other. Standard twisted-pair cables are wired straight through, meaning that the transmit contacts at one end of a cable are connected to the transmit contacts at the other end of that cable and the receive contacts to the receive contacts. To connect two computers directly using a twisted-pair cable and no hub, you must use a crossover cable in which the crossover is implemented in the cable wiring.
crosstalk A type of signal interference caused by signals transmitted on one pair of wires bleeding over into the other pairs. Crosstalk can cause network signals to degrade, eventually rendering them unusable. The individual wire pairs inside a twisted-pair cable are twisted at different rates because this helps to suppress the effects of crosstalk. Crosstalk is also the main reason you should not run other signals over the two unused wire pairs in a UTP Ethernet cable.
CSMA/CA See Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA)
CSMA/CD See Carrier Sense Multiple Access with Collision Detection (CSMA/CD).
cyclical redundancy check (CRC) An error detection mechanism in which a computer performs a calculation on a data sample with a specific algorithm and then transmits the data and the results of the calculation to another computer. The receiving computer then performs the same calculation and compares its results to those supplied by the sender. If the results match, the data has been transmitted successfully. If the results do not match, the data has been damaged in transit.
daemon The UNIX term for a computer program or process that runs continuously in the background and performs tasks at predetermined intervals or in response to specific events. Daemons typically perform server tasks, such as spooling print jobs, handling e-mail, and transmitting Web files. A daemon is called a service by Microsoft Windows operating systems.
DAT See digital audio tape (DAT).
data encapsulation The process by which information generated by an application is packaged for transmission over a network by successive protocols operating at the various layers of the Open Systems Interconnection (OSI) reference model. A protocol packages the data it receives from the layer above by adding a header (and sometimes a footer) containing protocol-specific information used to ensure that the data arrives at its destination intact.
datagram A term for the unit of data used by the Internet Protocol (IP) and other network layer protocols. Network layer protocols accept data from transport layer protocols and package it into datagrams by adding their own protocol headers. The protocol then passes the datagrams down to a data-link layer protocol for further packaging before they are transmitted over the network.
Datagram Delivery Protocol (DDP) The network layer protocol used by the AppleTalk protocol suite to carry end-to-end data across a network. See also AppleTalk.
data-link layer The second layer from the bottom of the Open Systems Interconnection (OSI) reference model. Protocols operating at the data-link layer are responsible for packaging network layer data, addressing it to its next destination, and transmitting it over the network. Some of the local area network (LAN) protocols operating at the data-link layer are Ethernet, Token Ring, and the Fiber Distributed Data Interface (FDDI). Wide area network (WAN) protocols operating at the data-link layer include the Point-to-Point Protocol (PPP) and the Serial Line Internet Protocol (SLIP).
D channel A digital communications channel running at 16 or 64 Kbps that is one of the fundamental units of service provided by the Integrated Services Digital Network (ISDN). D channels carry control traffic only, and are not factored into the user bandwidth provided by the service. The Basic Rate Interface (BRI) ISDN service consists of two B channels plus one 16-Kbps D channel; the Primary Rate Interface (PRI) service consists of 23 B channels and one 64-Kbps D channel. See also Integrated Services Digital Network (ISDN).
default gateway The router on the local network used by a TCP/IP client computer to transmit messages to computers on other networks. To communicate with other networks, TCP/IP computers consult their routing tables for the address of the destination network. If they locate the address, they send their packets to the router specified in the table entry, which relays them to the desired network. If no specific entry for the network exists, the computer sends the packets to the router specified in the default gateway entry, which the user (or a DHCP server) supplies as one of the basic configuration parameters of the TCP/IP client.
Destination Address A 48-bit field in data-link layer protocol headers that contains a hexadecimal sequence used to identify the network interface to which a frame will be transmitted.
Destination IP Address A 32-bit field in the Internet Protocol (IP) header that contains a value used to identify the network interface to which a packet will be transmitted.
DHCP See Dynamic Host Configuration Protocol (DHCP).
differential backup A type of backup job that uses a filter that causes it to back up only the files that have changed since the last full backup job. The filter evaluates the state of each file’s Archive bit, which a full backup job clears. Creating or modifying a file sets its Archive bit, and the differential job backs up only the files that have their Archive bit set. The differential job does not modify the state of the bits, so the next differential job will also back up all of the files that have changed since the last full backup. Differential jobs use more tape or other media than incremental jobs, because they repeatedly back up the same files, but they’re easier to restore in the event of a disaster. You only have to restore the last full backup and the most recent differential to completely restore a drive. See also incremental backup.
Differential Manchester encoding A physical layer encoding scheme, used on Token Ring networks, that is a variation on Manchester encoding.
digital audio tape (DAT) A data storage medium that uses cartridges containing 4-mm wide magnetic tape, most commonly for system backups.
digital linear tape (DLT) A data storage medium that uses cartridges containing one-half inch magnetic tape, most commonly used for system backups.
Digital Subscriber Line (DSL) A type of point-to-point, digital WAN connection that uses standard telephone lines to provide high-speed communications. DSL is available in many different forms, including Asymmetrical Digital Subscriber Line (ADSL) and High-bit-rate Digital Subscriber Line (HDSL). The various DSL technologies differ greatly in their speeds and in the maximum possible distance between the installation site and the telephone company’s nearest central office. DSL connections are used for many applications, ranging from LAN and PBX interconnections to consumer Internet access. See also Asymmetrical Digital Subscriber Line (ADSL).
Digital Subscriber Line Access Multiplexer (DSLAM) The hardware device located at the server side of an ADSL connection. See also ADSL Termination Unit-Remote (ATU-R), Asymmetrical Digital Subscriber Line (ADSL).
directory service A database containing information about network entities and resources, used as a guide to the network and an authentication resource by multiple users. Early network operating systems included basic flat file directory services, such as Microsoft Windows NT domains and the Novell NetWare bindery. Today’s directory services, such as Microsoft Active Directory directory service and Novell Directory Services (NDS) tend to be hierarchical and designed to support large enterprise networks. See also Active Directory directory service, Novell Directory Services (NDS).
direct route An Internet Protocol (IP) transmission to a destination on the local network, in which the Destination IP Address and the data-link layer protocol’s Destination Address identify the same computer. See also indirect route, in which the IP destination is on another network and the data-link layer Destination Address identifies a router on the local network used to access the destination network.
distance vector protocol A dynamic routing protocol that rates the relative efficiency of network routes by the number of hops to the destination. This is not necessarily an efficient method, because having networks of different speeds can cause a route with fewer hops to take longer to transmit data than one requiring more hops. The most common of the distance vector routing protocols is the Routing Information Protocol (RIP). See also link state protocol and Routing Information Protocol (RIP).
DIX Ethernet An acronym for Digital Equipment Corporation (DEC), Intel, and Xerox, the three corporations responsible for developing and publishing the original Ethernet standard.
DLT See digital linear tape (DLT).
DNS See Domain Name System (DNS).
domain A group of computers and other devices on a network that are administered as a single unit. On the Internet, domain names are hierarchical constructions (such as microsoft.com) that form the basis for the Domain Name System (DNS). On a Microsoft Windows network, a domain is a group of users, computers, and other resources for which information is stored in a directory service, on a server called a domain controller.
domain controller A computer running Microsoft Windows that has been designated for storing and processing directory service information. Microsoft Windows NT domains and the Active Directory directory service store their directory service databases on domain controllers, which also authenticate users accessing network resources.
Domain Name System (DNS) A distributed, hierarchical namespace designed to provide Transmission Control Protocol/Internet Protocol (TCP/IP) networks (such as the Internet) with friendly names for computers and users. Although TCP/IP computers use Internet Protocol (IP) addresses to identify each other, people work better with names. Domain Name System (DNS) provides a naming system for network resources and a service for resolving those names into IP addresses. TCP/IP computers frequently access DNS servers to send them the names of the computers they want to access. The DNS server communicates with other DNS servers on the network to find out the IP address associated with the requested name and then sends it back to the client computer, which initiates communications with the destination system using its IP address.
double ring A network cabling topology that consists of two separate rings with traffic running in opposite directions, used primarily by the Fiber Distributed Data Interface (FDDI) protocol. Devices are connected to both rings, providing a measure of fault tolerance in the event of a cable failure that causes a break in one of the rings. Unlike the standard ring topology, the double ring is usually implemented physically, not logically. See also ring, topology, Fiber Distributed Data Interface (FDDI).
driver A software component that enables an application or operating system to use a particular hardware device. Also called a device driver.
drive spanning A process by which a computer creates a single logical storage unit called a volume by combining the disk space of two or more drives. The volume appears to users as a single logical entity, but data is actually being stored on multiple drives. The primary drawback of this arrangement is that if one of the drives fails, the entire volume is lost.
DSL See Digital Subscriber Line (DSL).
DSLAM See Digital Subscriber Line Access Multiplexer (DSLAM).
DSL modem Inaccurate terminology for the hardware unit that provides ADSL client connectivity, which is correctly called an ADSL Termination Unit-Remote (ATU-R).
duplexing A data availability technique that involves storing identical copies of data on two different drives connected to different host adapters. The drives appear as a single volume to users, and all files written to the volume are copied to both drives automatically. If one of the drives or adapters fails, the other continues to make the data available until the failed component is repaired or replaced. See also mirroring.
dynamic allocation An operational mode of Dynamic Host Configuration Protocol (DHCP) servers in which the server assigns an Internet Protocol (IP) address and other TCP/IP configuration settings to a client from a pool of addresses, and then reclaims them when a lease of a given duration expires. This enables you to move computers to different subnets without having to manually release the previously allocated IP addresses from the other subnets. See also automatic allocation, manual allocation, Dynamic Host Configuration Protocol (DHCP).
Dynamic Host Configuration Protocol (DHCP) A service that automatically configures the Transmission Control Protocol/Internet Protocol (TCP/IP) client computers on a network by assigning them unique Internet Protocol (IP) addresses and other configuration parameters. DHCP servers can assign IP addresses to clients from a pool and reclaim them when a lease of a set duration expires. Virtually all operating systems include a DHCP client, and most of the major server operating systems—such as Microsoft Windows 2000 Server, Microsoft Windows NT Server, Novell NetWare, and many forms of UNIX—include DHCP server software. DHCP is a cross-platform service that can support various operating systems with a single server. See also automatic allocation, dynamic allocation, manual allocation.
dynamic routing A system in which routers automatically build their own routing tables using specialized protocols to communicate with other nearby routers. By sharing information in this way, a router builds up a composite picture of the internetwork on which it resides, enabling it to route traffic more efficiently. The two basic types of routing protocols are distance vector routing protocols, like the Routing Information Protocol (RIP), and link state routing protocols, like the Open Shortest Path First (OSPF) protocol. See also Routing Information Protocol (RIP).
E-1 A dedicated telephone connection, also called a leased line, running at 2.048 Mbps. An E-1 is the European equivalent of a T-1. See also T-1, leased line.
E-3 A dedicated telephone connection, also called a leased line, running at 34.368 Mbps. An E-3 is the European equivalent of a T-3. See also T-3, leased line.
EIA/TIA See Telecommunications Industry Association/Electronic Industries Alliance (TIA/EIA)
electromagnetic interference (EMI) In data networking, any electromagnetic disturbance that interrupts or degrades the transmission of signals over a network medium.
e-mail A service that transmits messages in electronic form to specific users on a network.
EMI See electromagnetic interference.
encryption The process of making information indecipherable in order to protect it from unauthorized viewing or use, especially during transmission or when the data is stored on a transportable magnetic medium. A key is required to decode the information.
end system On a Transmission Control Protocol/Internet Protocol (TCP/IP) network, a computer or other device that is the original sender or ultimate recipient of a transmission. The end systems in a TCP/IP transmission are identified by the Source IP Address and Destination IP Address fields in the Internet Protocol (IP) header. All of the other systems (that is, routers) involved in the transmission are known as intermediate systems.
ephemeral port number A Transmission Control Protocol (TCP) or User Datagram Protocol (UDP) port number of 1024 or higher, chosen at random by a Transmission Control Protocol/Internet Protocol (TCP/IP) client computer during the initiation of a transaction with a server. Because the client initiates the communication with the server, it can use any port number beyond the range of the well-known port numbers (which run up to 1023). The server reads the ephemeral port number from the transport layer protocol header’s Source Port field and uses it to address its replies to the client. See also well-known port.
Ethernet A common term used to describe Institute of Electrical and Electronics Engineers (IEEE) 802.3, a data-link layer local area network (LAN) protocol developed in the 1970s, which is now the most popular protocol of its kind in the world. Ethernet runs at 10 Mbps, is based on the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Media Access Control (MAC) mechanism, and supports a variety of physical layer options, including coaxial, unshielded twisted pair (UTP), and fiber optic cables. More recent revisions of the protocol support speeds of 100 Mbps (Fast Ethernet) and 1000 Mbps (Gigabit Ethernet). See also Carrier Sense Multiple Access with Collision Detection (CSMA/CD).
Fast Ethernet The updated version of the Ethernet local area network (LAN) protocol that increases transmission speed from 10 to 100 Mbps, preserving nearly all of Ethernet’s defining elements, such as its frame format, its physical layer options, and the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Media Access Control (MAC) mechanism. Fast Ethernet is defined in a new document, Institute of Electrical and Electronics Engineers (IEEE) 802.3u, published in 1995. Fast Ethernet supports three primary physical layer options: 100Base-TX for Category 5 (CAT5) unshielded twisted pair (UTP) cable, 100Base-T4 for Category 3 (CAT3) UTP cable, and 100Base-FX for multimode fiber optic cable.
fast link pulse (FLP) The signal generated by Fast Ethernet network interface adapters and hubs, which the devices use to signal that they have been cabled together properly and to automatically negotiate the fastest transmission speed they have in common. When an adapter or hub receives the FLP signal from the device to which it is connected, it activates a light-emitting diode (LED), which indicates that communication is taking place. FLP signals are completely compatible with the normal link pulse (NLP) signals used by 10Base-T Ethernet devices, differing only in that they include a link code word that specifies the transmission speeds they support.
FAT See file allocation table (FAT).
FDDI See Fiber Distributed Data Interface (FDDI).
Fiber Distributed Data Interface (FDDI) A data-link layer local area network (LAN) protocol running at 100 Mbps, designed for use with fiber optic cable. Typically used for backbone networks, FDDI uses the token passing Media Access Control (MAC) mechanism and supports a double ring topology that provides fault tolerance in the event of a system disconnection or cable failure. Originally the principal 100-Mbps LAN protocol, FDDI has since largely been replaced by the Fast Ethernet and Gigabit Ethernet fiber optic options.
fiber optic A network cable technology that uses signals consisting of pulses of light rather than the electrical charges used by copper cables. Fiber optic cable is completely resistant to electromagnetic interference (EMI) and is also able to span far longer distances than copper cables, indoors or outdoors. The core conductors in a fiber optic cable are made of glass or plastic and are surrounded by a cladding that reflects the light back on itself, keeping it in the core of the cable. The light source is a light-emitting diode (LED) or a laser, depending on the type of cable. Fiber optic cable is generally more efficient than copper-based cable in almost every way, but it’s more expensive than copper and more difficult to install, requiring specialized tools and skills. See also multimode fiber, singlemode fiber.
Fiber Optic Inter-Repeater Link (FOIRL) The earliest Ethernet physical layer specification to use fiber optic cable. FOIRL was defined in the DIX Ethernet II document. FOIRL uses 62.5/125 multimode fiber optic cable in a star topology, with a maximum segment length of 1000 meters. FOIRL was rarely used, and was replaced in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard by the 10Base-F specification (10Base-FL, 10Base-FB, and 10Base-FP).
file allocation table (FAT) The file system used by the DOS operating system, which is based on a table that specifies which disk clusters contain the files stored on a disk. The Microsoft Windows 95, Microsoft Windows 98, Microsoft Windows Me, Microsoft Windows NT, and Microsoft Windows 2000 operating systems currently support the 16-bit version of the FAT file system. Windows 95 OSR2, Windows 98, Windows Me, and Windows 2000 also support FAT-32, a newer version that uses 32-bit FAT entries, enabling the file system to support much larger disk drives. The FAT file system is sufficient for a standard workstation but lacks the security capabilities required by server drives. For this reason, the Microsoft operating systems designed for heavier network use—Microsoft Windows 2000 and Microsoft Windows NT—also include the NT file system (NTFS), which has greater security capabilities.
File Transfer Protocol (FTP) An application layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol designed to perform file transfers and basic file management tasks on remote computers. FTP is a mainstay of Internet communications. FTP client support is integrated into most Web browsers, and FTP server support is integrated into many Web server products. FTP is also an important UNIX tool—all UNIX systems support both FTP client and server functions. FTP is unique among TCP/IP protocols because it uses two simultaneous Transmission Control Protocol (TCP) connections. One, a control connection, remains open during the entire life of the session between the FTP client and the FTP server. When the client initiates a file transfer, a second connection is opened between the two computers to carry the transferred data. This connection closes at the conclusion of the data transfer.
firewall A hardware or software product designed to isolate part of an internetwork to protect it against intrusion by outside processes. Typically used to protect a private network from intrusion from the Internet, firewalls use a number of techniques to provide this protection, while still allowing certain types of traffic through. Some of these techniques include packet filtering and Network Address Translation (NAT). Firewalls were once intended only for large network installations, but there are now smaller firewall products designed to protect small networks and individual computers from Internet intruders.
fish tape A tool used by cable installers to push or pull cables up or down inside walls. It consists of a flexible metal tape with a hook on the end wound onto a reel (much like a plumber’s snake). Cable installers connect the end of a cable to the hook and draw it through a wall by unwinding a length of tape and extending it through the cavity inside the wall.
flow control A function of certain data transfer protocols that enables a system receiving data to transmit signals to the sender instructing it to slow down or speed up its transmissions. This prevents the receiving system from overflowing its buffers and being forced to discard incoming data. For example, the Transmission Control Protocol (TCP) implements its flow control mechanism by using a Windows field to specify the number of bytes that it can receive from the sender.
FLP See fast link pulse (FLP).
FOIRL See Fiber Optic Inter-Repeater Link (FOIRL).
fox and hound wire tester The colloquial name for a simple type of cable tester, also called a tone generator and locator.
frame A unit of data that is constructed, transmitted, and received by data-link layer protocols such as Ethernet and Token Ring. Data-link layer protocols create frames by packaging the data they receive from network layer protocols inside a header and footer. Frames can be different sizes, depending on the protocol used to create them.
frame relay A wide area networking technology in which two systems are each connected to a frame relay network called a cloud, and a virtual circuit is established between them through the cloud. The advantages of frame relay over a leased line are that the amount of bandwidth provided by the connection is flexible and that one site can be connected to numerous other sites using multiple virtual circuits. See also leased line.
FTP See File Transfer Protocol (FTP).
full-duplex A form of communications in which two connected systems can send signals to the other system simultaneously. For example, a telephone call (in which both parties can talk at once at any time) is an example of full-duplex communication, while a citizen’s band (CB) radio (on which you must press a key to transmit signals and release the key to receive them) is an example of a half-duplex communication device.
gateway On a Transmission Control Protocol/Internet Protocol (TCP/IP) network, the term “gateway” is often used synonymously with the term “router,” referring to a network layer device that connects two networks together and relays traffic between them as needed, such as the default gateway specified in a TCP/IP client configuration. However, the term “gateway” is also used to refer to an application layer device that relays data between two different services, such as an e-mail gateway that enables two separate e-mail services to communicate with each other.
GB Gigabyte, equal to 1000 megabytes (MB) or 1,000,000 kilobytes (KB) or 1,000,000,000 bytes.
GBps See gigabytes per second.
Gbps See gigabits per second.
gif A compressed file format commonly used to store graphic images in bitmap form.
Gigabit Ethernet The latest version of the Ethernet data-link layer protocol, defined in the Institute of Electrical and Electronics Engineers (IEEE) 802.3z and IEEE 802.3ab documents and running at 1000 Mbps. Gigabit Ethernet is designed for backbone networks and server connections and supports a variety of unshielded twisted pair (UTP) and fiber optic cabling options. The UTP option uses all four of the wire pairs in the cable to carry signals, instead of the two pairs used by most of the other Ethernet types. As with the other Ethernet varieties, Gigabit Ethernet uses the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Media Access Control (MAC) mechanism.
gigabits per second (Gbps) A unit of measurement typically used to measure network transmission speed.
gigabytes per second (GBps) A unit of measurement typically used to measure the speed of data storage devices.
grandfather-father-son A media rotation scheme used by many backup software programs. “Grandfather” refers to monthly backup jobs, “father” to weekly jobs, and “son” to daily jobs.
graphical user interface (GUI) An element of a program or operating system that takes advantage of the computer’s graphical capabilities, enabling the user to manipulate software components represented by objects on the display.
half-duplex A form of communications in which two connected systems can only send signals in one direction at a time. For example, a citizen’s band (CB) radio (on which you must press a key to transmit signals and release the key to receive them) is an example of a half-duplex communications device, whereas a telephone call (in which both parties can talk at once at any time) is an example of full-duplex communication. Most local area network (LAN) protocols operate in half-duplex mode, although there is a full-duplex version of Ethernet.
HDSL See High-bit-rate Digital Subscriber Line (HDSL).
hierarchical star A network cabling topology in which a standard star network is augmented by the addition of one or more hubs, connected to the original ones. Also called a branching tree network. See also topology.
High-bit-rate Digital Subscriber Line (HDSL) A point-to-point, digital wide area network (WAN) technology used by telephone companies and other large corporations to transmit data at T-1 speeds.
hop A unit of measurement used to quantify the length of a route between two computers on an internetwork, as indicated by the number of routers that packets must pass through to reach the destination end system. For example, if packets must be forwarded by four routers in the course of their journey from end system to end system, the destination is said to be four hops away from the source. Distance vector routing protocols like the Routing Information Protocol (RIP) use the number of hops as a means to compare the relative efficiency of routes.
Hosts An American Standard Code for Information Interchange (ASCII) text file used by Transmission Control Protocol/Internet Protocol (TCP/IP) computers to resolve host names into Internet Protocol (IP) addresses. The Hosts file is a simple list of the host names used by TCP/IP computers and their equivalent IP addresses. When a user or an application refers to a computer using a host name, the TCP/IP client looks it up in the Hosts file to determine its IP address. The Hosts file was the original name resolution method for what later became the Internet, until the number of computers on the network grew too large to manage using this technique. Eventually, the Domain Name System (DNS) was created to perform the same function in a more efficient and manageable way. TCP/IP computers can still use a Hosts file for name resolution, but because the names and addresses of each computer must be added manually, this method is rarely used today.
HTTP See Hypertext Transfer Protocol (HTTP).
hub A hardware component to which cables running from computers and other devices are connected, joining all of the devices into a network. In most cases, the term “hub” refers to an Ethernet multiport repeater, a device that amplifies the signals received from each connected device and forwards them to all of the other devices simultaneously. See also multiport repeater.
Hypertext Transfer Protocol (HTTP) An application layer protocol that is the basis for World Wide Web communications. Web browsers generate HTTP Get request messages containing Uniform Resource Locators (URLs) and transmit them to Web servers, which reply with one or more HTTP Response messages containing the requested files. HTTP traffic is encapsulated using the Transmission Control Protocol (TCP) at the transport layer and the Internet Protocol (IP) at the network layer. Each HTTP transaction requires a separate TCP connection.
IANA See Internet Assigned Numbers Authority (IANA).
IBM data connector (IDC) A proprietary connector used to attach Token Ring systems to multistation access units (MAUs) using Type 1 cables and to connect MAUs together. On today’s Token Ring networks, Type 1 cables and IDC connectors have largely been replaced by RJ-45 connectors and unshielded twisted pair (UTP) cables.
ICMP See Internet Control Message Protocol (ICMP).
IDC See IBM data connector (IDC).
IEEE See Institute of Electrical and Electronic Engineers (IEEE).
IEEE 802.2 A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the Logical Link Control (LLC) sublayer used by the IEEE 802.3, IEEE 802.5, and other protocols.
IEEE 802.3 A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining what is commonly referred to as the Ethernet protocol. Although there are slight differences from the original DIX Ethernet standards, such as the omission of the Ethertype field and the separation of the data-link layer into two sublayers—the Media Access Control (MAC) sublayer and the Logical Link Control (LLC) sublayer—IEEE 802.3 retains the defining characteristics of Ethernet, including the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) MAC mechanism. IEEE 802.3 also adds to the physical layer options defined in the DIX Ethernet standards by including support for unshielded twisted pair (UTP) cable.
IEEE 802.3ab A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining an implementation of the 1000-Mbps Gigabit Ethernet protocol using Category 5 (CAT5) unshielded twisted pair (UTP) cable and a 100-meter maximum segment length. Released after the original Gigabit Ethernet protocol standard (IEEE 802.3z), this specification is intended to be an upgrade path to Gigabit Ethernet for existing UTP regular or Fast Ethernet networks. To achieve a transmission speed of 1000 Mbps, this standard calls for the use of all four pairs of wires in the cable, plus a signaling scheme called Pulse Amplitude Modulation-5 (PAM-5).
IEEE 802.3d A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the Fiber Optic Inter-Repeater Link (FOIRL) specification.
IEEE 802.3i A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the 10Base-T specification.
IEEE 802.3j A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the 10Base-FP, 10Base-FB, and 10Base-FL specifications.
IEEE 802.3u A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the Fast Ethernet data-link layer local area network (LAN) protocol. Running at 100 Mbps, Fast Ethernet uses the same frame format and the Carrier Sense Multiple Access with Collision Detection (CSMA/ CD) Media Access Control (MAC) mechanism as standard Ethernet, and supports three physical layer options: 100Base-TX, 100Base-T4, and 100Base-FX. Many Fast Ethernet hardware products support both 10 Mbps and 100 Mbps speeds and use an enhanced link pulse signal called fast link pulse (FLP) to negotiate the fastest possible transmission speed with the connected device.
IEEE 802.3z A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining the 1000-Mbps Gigabit Ethernet data-link layer protocol. Designed primarily for use on backbone networks and server connections that require high speeds, IEEE 802.3z was the first Gigabit Ethernet standard published, and includes a variety of physical layer options, most of which call for various types of fiber optic cable. Like the other varieties of Ethernet, Gigabit Ethernet uses the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Media Access Control (MAC) mechanism.
IEEE 802.5 A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining a Token Ring-like data-link layer protocol. See also Token Ring.
IEEE 802.11 A standards document published by the Institute of Electrical and Electronic Engineers (IEEE) defining a wireless local area network (LAN) running at speeds of up to 11 Mbps using any one of three physical layer technologies: direct sequence spread spectrum (DSSS), frequency hopping spread spectrum (FHSS), and infrared.
IETF See Internet Engineering Task Force (IETF).
ifconfig A UNIX utility program used to configure a network interface and display the network interface’s configuration parameters. The similar Ipconfig.exe is a program available in Microsoft Windows that performs the display functions only.
IMAP See Internet Mail Access Protocol (IMAP).
incremental backup A type of backup job that uses a filter that causes it to back up only the files that have changed since the last backup job. The filter evaluates the state of each file’s Archive bit, which a full backup job or an incremental backup job clears. Creating or modifying a file sets its Archive bit, and the incremental job backs up only the files with an Archive bit that is set. It then resets the Archive bits (unlike a differential job, which does not reset the bits). Incremental jobs use the least amount of tape or other medium, but they are more difficult to restore in the event of a disaster. You must restore the last full backup job and all of the incremental jobs performed since that last full backup, in the correct chronological order, to fully restore a drive. See also differential backup.
Independent Computing Architecture (ICA) A protocol developed by Cyrix Systems that provides communication between thin clients and network servers. Thin clients are terminals that exchange keystrokes, mouse actions, and display data with servers that run the user operating system and applications.
indirect route An Internet Protocol (IP) transmission to a destination on a different network, in which the Destination IP Address and the data-link layer protocol’s Destination Address identify different computers. See also direct route, in which the IP destination is on the same network and the data-link layer Destination Address identifies the same computer as the Destination IP Address.
infrastructure topology A type of communication used on wireless LANs (WLANs) in which devices equipped with wireless network interface adapters communicate with a standard cabled network using a network access point. See also ad hoc topology, network access point.
Institute of Electrical and Electronics Engineers (IEEE) An organization, founded in 1894, dedicated to the development and publication of standards for the computer and electronics industries. IEEE is best known in computer networking for the IEEE 802 series of documents defining the data-link layer local area network (LAN) protocols commonly known as Ethernet and Token Ring.
Integrated Services Digital Network (ISDN) A dial-up communications service that uses standard telephone lines to provide high-speed digital communications. Originally conceived as a replacement for the existing analog telephone service, it never achieved its anticipated popularity. Today, ISDN is used in the United States primarily as an Internet access technology, although it is more commonly used for wide area network (WAN) connections in Europe and Japan. The two most common ISDN services are the Basic Rate Interface (BRI), which provides two 64-Kbps B channels and one 16-Kbps D (control) channel, and the Primary Rate Interface (PRI), which provides 23 64-Kbps B channels and one 64-Kbps D channel.
intelligent hub Also called a smart hub, a local area network (LAN) cabling nexus that not only functions at the physical layer by propagating traffic to all of the other computers on the network, but can buffer data and retransmit it out through specific ports as needed, and in some cases monitor the activity on all of its ports and transmit information about its status to a network management console.
intermediate system On a Transmission Control Protocol/ Internet Protocol (TCP/IP) network, a router that relays traffic generated by an end system from one network to another. The end systems in a TCP/IP transmission are identified by the Source IP Address and Destination IP Address fields in the Internet Protocol (IP) header. All of the other systems (that is, routers) involved in the transmission are known as intermediate systems.
International Organization for Standardization (ISO) An organization, founded in 1946, that consists of standards bodies from over 75 countries, such as the American National Standards Institute (ANSI) from the United States. The ISO is responsible for the publication of many computer-related standards, the most well-known of which is “The Basic Reference Model for Open Systems Interconnection,” commonly known as the Open Systems Interconnection (OSI) reference model. (ISO is not merely an acronym; it’s a name derived from the Greek word isos, meaning “equal.”)
International Telecommunications Union (ITU) An organization, founded in 1865, devoted to the development of treaties, regulations, and standards governing telecommunications. Since 1992, it has included the standards development organization formerly known as the Comité Consultatif International Téléphonique et Télégraphique (CCITT), which was responsible for the creation of modem communication, compression, and error correction standards.
Internet A packet-switching internetwork that consists of thousands of individual networks and millions of computers located around the world. The Internet is not owned or administered by any central managing body; all administration chores are distributed among users all over the network.
internet See internetwork.
Internet Assigned Numbers Authority (IANA) The organization responsible for assigning unique parameter values for the Transmission Control Protocol/Internet Protocol (TCP/IP) protocols, including Internet Protocol (IP) address assignments for networks and protocol number assignments. The “Assigned Numbers” Requests for Comments (RFC) document (currently RFC 1700) lists all of the protocol number assignments and many other unique parameters regulated by the IANA.
Internet Control Message Protocol (ICMP) A network layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol that carries administrative messages, particularly error messages and informational queries. ICMP error messages are primarily generated by intermediate systems that have no other means of signaling errors to the end system that transmitted the packet because the packets they route travel no higher than the network layer. Typical ICMP error messages inform the sender that the network or host to which a packet is addressed could not be found, or that the Time To Live value for a packet has expired. ICMP query messages request information (or simply a response) from other computers, and are the basis for TCP/IP utilities like Ping, which is used to test the ability of one computer on a network to communicate with another.
Internet Engineering Task Force (IETF) The primary standards ratification body for the Transmission Control Protocol/Internet Protocol (TCP/IP) protocol and the Internet. The IETF publishes Requests for Comments (RFCs), which are the working documents for what eventually become Internet standards. The IETF is an international body of network designers, operators, software programmers, and other technicians, all of whom devote part of their time to the development of Internet protocols and technologies.
Internet Mail Access Protocol (IMAP) An application layer Transmission Control Protocol/ Internet Protocol (TCP/IP) protocol used by e-mail clients to download mail messages from a server. E-mail traffic between servers and outgoing e-mail traffic from clients to servers uses the Simple Mail Transfer Protocol (SMTP). See also Post Office Protocol 3 (POP3).
Internet Protocol (IP) The primary network layer protocol in the Transmission Control Protocol/ Internet Protocol (TCP/IP) suite. IP is the protocol that is ultimately responsible for end-to-end communications on a TCP/IP internetwork, and includes functions such as addressing, routing, and fragmentation. IP packages data that it receives from transport layer protocols into data units called datagrams by applying a header containing the information needed to transmit the data to its destination. The IP addressing system uses 32-bit addresses to uniquely identify the computers on a network, and specifies the address of the destination system as part of the IP header. IP is also responsible for routing packets to their destinations on other networks by forwarding them to other routers on the network. When a datagram is too large to be transmitted over a particular network, IP breaks it into fragments and transmits each in a separate packet.
Internet service provider (ISP) A type of company whose business is supplying consumers or businesses with Internet access. At the consumer level, an ISP provides users with dial-up access to the ISP’s networks, which are connected to the Internet, as well as other end-user services, such as access to Domain Name System (DNS), e-mail, and news servers. At the business level, ISPs provide high-bandwidth Internet connections using leased telephone lines or other technologies, and sometimes also provide other services, such as registered Internet Protocol (IP) addresses, Web site hosting, and DNS domain hosting.
internetwork A group of interconnected local area networks (LANs) and/or wide area networks (WANs) that are connected so that any computer can transmit data to any other computer. The networks are connected by routers, which are responsible for relaying packets from one network to another. The largest example of an internetwork is the Internet, which is composed of thousands of networks located around the world. Private internetworks consist of a smaller number of LANs, often at various locations and connected by WAN links.
Internetwork Packet Exchange (IPX) A network layer protocol used by Novell NetWare networks. IPX performs many of the same functions as the Internet Protocol (IP), but instead of being a self-contained addressing system like IP, IPX is designed for use on local area networks (LANs) only and uses a network identifier assigned by the network administrator plus the network interface adapter’s hardware address to identify the individual computers on the network. Unlike IP, IPX is not based on an open standard. Novell owns all rights to the protocols of the IPX protocol suite, although Microsoft has developed its own IPX-compatible protocol for inclusion in the Windows operating systems.
intranet A Transmission Control Protocol/Internet Protocol (TCP/IP) network owned by a private organization that provides services such as Web sites only to that organization’s users.
IP See Internet Protocol (IP).
IP address A 32-bit address assigned to Transmission Control Protocol/Internet Protocol (TCP/IP) client computers and other network equipment that uniquely identifies that device on the network. The Internet Protocol (IP) uses IP addresses to transmit packets to the destinations. Expressed as four 8-bit decimal values separated by periods (for example, 192.168.71.19), the IP address consists of a network identifier (which specifies the network that the device is located on) and a host identifier (which identifies the particular device on that network). The sizes of the network and host identifiers can vary depending on the address class. For a computer to be accessible from the Internet, it must have an IP address containing a network identifier registered with the Internet Assigned Numbers Authority (IANA).
IP Authentication Header (AH) An IPSec protocol that provides data integrity and anti-replay functions, but not encryption.
Ipconfig.exe A Microsoft Windows 2000 and Microsoft Windows NT command-line utility used to view the Transmission Control Protocol/Internet Protocol (TCP/IP) configuration parameters for a particular computer. A graphical version of the tool, Winipcfg.exe, is included with Microsoft Windows 95, Microsoft Windows 98, and Microsoft Windows Me. Ipconfig.exe is most useful on computers with TCP/IP clients configured automatically by a Dynamic Host Configuration (DHCP) server because it is the easiest way to view the assigned settings for the client system. You can also use Ipconfig.exe to release and renew DHCP-assigned TCP/IP configuration parameters.
IP Encapsulating Security Payload (ESP) An IPSec protocol that provides encryption, data integrity, and anti-replay functions.
IPSec See IP Security protocol (IPSec).
IP Security protocol (IPSec) A set of Transmission Control Protocol/Internet Protocol (TCP/IP) protocols designed to provide encrypted network layer communications. For computers to communicate using IPSec, they must share a public key.
IPv6 A new version of the Internet Protocol (IP) that expands the IP address space from 32 to 128 bits. See also Internet Protocol (IP).
IPX See Internetwork Packet Exchange (IPX).
ISDN See Integrated Services Digital Network (ISDN).
ISO See International Organization for Standardization (ISO).
ISP See Internet service provider (ISP).
ITU See International Telecommunications Union (ITU).
Jaz The proprietary name for a magnetic cartridge drive holding 1 or 2 gigabytes (GB) of data.
JPEG jpg A compressed file format commonly used to store graphic images in bitmap form.
Kbps See kilobits per second.
Kerberos An authentication protocol that uses public key technology to provide users with secured access to network resources.
kilobits per second (Kbps) A unit of measurement typically used to measure network transmission speed.
LAN See local area network (LAN).
late collision On an Ethernet network, a data collision between two transmitted packets that occurs after one or both packets has completely left the transmitting system. The physical layer specifications of the Ethernet protocols are designed to ensure that the first bit transmitted by a computer reaches its destination before the last bit leaves that computer. This allows the transmitting system to detect collisions when they occur. Collisions are normal on an Ethernet network, but if a cable segment is too long, or if there are too many hubs on the path to the destination, late collisions can occur after packets have left the transmitting system, which makes it impossible for the Ethernet adapter in the transmitting system to detect them. Unlike the normal type of collision, late collisions are a serious problem on an Ethernet network and should be addressed immediately. See also collision, Ethernet.
Layer 2 Tunneling Protocol (L2TP) A protocol used to establish virtual private network (VPN) connections across the Internet. See also virtual private network (VPN).
leased line A permanent telephone connection between two points that provides a predetermined amount of bandwidth at all times. See also T-1, T-3.
lease identification cookie A string that consists of a computer’s Internet Protocol (IP) address and its hardware address, which a Dynamic Host Configuration Protocol (DHCP) server uses to uniquely identify a client in its database. See also Dynamic Host Configuration Protocol (DHCP).
linear tape-open (LTO) A data storage medium that uses cartridges containing one-half-inch wide magnetic tape, most commonly used for system backups.
link code word A 16-bit data packet included in the fast link pulse (FLP) signals generated by Fast Ethernet devices that contains the speeds at which the device can transmit data and whether or not the device supports full-duplex transmissions.
link pulse A signal transmitted by Ethernet devices that indicates when the devices are communicating properly. Ethernet unshielded twisted pair (UTP) network interface adapters and hubs typically have light-emitting diodes (LEDs) that light up when the device receives a link pulse signal from a device to which it is connected. 10Base-T devices use a normal link pulse (NLP) signal, which is used only for link integrity testing. Fast Ethernet devices use a fast link pulse (FLP) signal, which also includes a link code word that enables the devices to negotiate the fastest possible transmission speed they have in common. See also fast link pulse (FLP), normal link pulse (NLP).
link segment A network segment that connects only two computers together, such as a cable that connects a computer to a hub. See also mixing segment, which consists of cables that run from computer to computer in daisy-chain fashion. The Ethernet protocol distinguishes between mixing segments and link segments in the physical layer configuration guidelines that specify how many repeaters are permitted on a network.
link-state protocol A dynamic routing protocol that rates the relative efficiency of network routes by the properties of the connections providing access to the destination. See also distance vector protocols, which use the number of hops to rate the efficiency of a network. The most common of the link state protocols is the Open Shortest Path First (OSPF) protocol. See also Open Shortest Path First (OSPF).
LLC See Logical Link Control (LLC) sublayer.
Lmhosts An American Standard Code for Information Interchange (ASCII) text file used by Windows Transmission Control Protocol/Internet Protocol (TCP/IP) computers to resolve Network Basic Input/Output System (NetBIOS) names into Internet Protocol (IP) addresses. Like the Hosts file used to resolve host names into IP addresses, an Lmhosts file is a list of the NetBIOS names assigned to computers on the network and their corresponding IP addresses. Lmhosts files can also contain special entries used to preload the computer’s NetBIOS name cache or to identify the domain controllers on the network. Windows systems can use individual Lmhosts files for NetBIOS name resolution, but they more commonly use either network broadcast transmissions or the Windows Internet Naming Service (WINS).
local area network (LAN) A collection of computers that are connected to each other using a shared medium. The computers communicate with each other using a common set of protocols. See also wide area network (WAN), metropolitan area network (MAN).
Logical Link Control (LLC) sublayer One of the two sublayers of the data-link layer defined by the Institute of Electrical and Electronic Engineers (IEEE) 802 standards. The LLC standard (IEEE 802.2) defines additional fields carried within the data field of data-link layer protocol headers. See also Media Access Control (MAC) sublayer.
loopback connector A hardware tool used to test a network interface adapter by redirecting outgoing signals back into the device.
LTO See linear tape-open (LTO).
MAC See Media Access Control (MAC).
MAN See metropolitan area network (MAN).
management information base (MIB) An object-oriented database in which a network management agent stores the information that it will eventually transmit to a network management console using a protocol like the Simple Network Management Protocol (SNMP). Agents are built into network hardware and software products to enable them to report the status of the product to a central console monitored by a network administrator.
Manchester encoding A self-timing physical layer encoding scheme used on Ethernet networks.
manual allocation An operational mode of Dynamic Host Configuration Protocol (DHCP) servers in which the server assigns clients Internet Protocol (IP) addresses and other TCP/IP configuration settings specified by the server administrator for each computer. The IP addresses are not assigned randomly from a pool, as in the automatic and dynamic allocation modes. The end result is no different than configuring the TCP/IP clients by hand, but using the manual allocation mode of a DHCP server prevents the administrator from having to travel to the client computer and prevents other computers on the network from being assigned duplicate addresses. Manual allocation is typically used for clients that must have a specific IP address, such as a Web server that must be accessible from the Internet using a DNS name. See also Dynamic Host Configuration Protocol (DHCP).
MAU See multistation access unit.
maximum transmission unit (MTU) The largest physical packet size that a system can transmit over a network. As packets are routed through an internetwork, they might have to pass through individual networks with different MTUs. When a packet exceeds the MTU for a particular network, the network layer protocol (Internet Protocol, or IP, in most cases) divides the packet into fragments smaller than the MTU for the outgoing network. The protocol then repackages each fragment into a separate packet and transmits them. If necessary, fragments can be split into still smaller fragments by other routers along the way to the destination. Packets remain fragmented for the rest of their journey and are not reassembled until they reach the end system that is the packet’s ultimate destination.
MB Megabyte, equal to 1000 kilobytes (MB) or 1,000,000 bytes.
MBps Megabytes per second, a unit of measurement typically used to measure the speed of data storage devices.
Mbps See megabits per second.
media In networking, a term used to describe the data-carrying hardware mechanism that computers and other network devices use to send information to each other. In computers, a term used to describe a means of storing data in a permanent fashion, such as a hard or floppy disk.
Media Access Control (MAC) A method by which computers determine when they can transmit data over a shared network medium. When multiple computers are connected to a single network segment, two computers transmitting data at the same time cause a collision, which destroys the data. The MAC mechanism implemented in the data-link layer protocol prevents these collisions from occurring or permits them to occur in a controlled manner. The MAC mechanism is the defining characteristic of a data-link layer local area network (LAN) protocol. The two most common MAC mechanisms in use today are Carrier Sense Multiple Access with Collision Detection (CSMA/CD), which is used by Ethernet networks, and token passing, which is used by Token Ring and Fiber Distributed Data Interface (FDDI) networks, among others.
Media Access Control (MAC) sublayer One of the two sublayers of the data-link layer defined by the Institute of Electrical and Electronic Engineers (IEEE) 802 standards. The MAC sublayer defines the mechanism used to regulate access to the network medium. See also Logical Link Control (LLC) sublayer.
megabytes per second (MBps) A unit of measurement typically used to measure the speed of data storage devices.
mesh In local area networking, a cable topology in which each device is connected to every other device with a separate length of cable. In this respect, the mesh network is purely theoretical, because it would be impractical to implement or impossible with more than a handful of devices. In internetworking, the term “mesh” refers to a fabric of connected networks that provides more than one route to a particular destination. See also topology.
metric A field in a Transmission Control Protocol/Internet Protocol (TCP/IP) computer’s routing table that contains a value rating the relative efficiency of a particular route. When routing packets, a router scans its routing table for the desired destination, and if there are two possible routes to that destination listed in the table, the router chooses the one with the lowest metric value. Depending on how the routing information is inserted into the table, the metric can represent the number of hops needed to reach the destination network, or it can contain a value that reflects the actual time needed to reach the destination.
metropolitan area network (MAN) A data network that services an area larger than a local area network (LAN) and smaller than a wide area network (WAN). Most MANs today service communities, towns, or cities and are operated by cable television (CATV) companies using fiber optic cable.
MIB See management information base (MIB).
minimal routing The process of routing Internet Protocol (IP) using only the default routing table entries created by the operating system. See also static routing, dynamic routing.
mirroring A data availability technique that involves storing identical copies of data on two different drives connected to a single host adapter. The drives appear as a single volume to users, and all files written to the volume are automatically copied to both drives. If one of the drives fails, the other continues to make the data available until the failed drive is repaired or replaced. See also duplexing.
mixing segment A network segment that connects more than two computers, such as a thin Ethernet segment, which consists of cables that run from computer to computer in daisy-chain fashion. The Ethernet protocol distinguishes between mixing segments and link segments in the physical layer configuration guidelines that specify how many repeaters are permitted on a network. See also link segment.
modem Short for modulator/demodulator, a hardware device that converts the digital signals generated by computers into analog signals suitable for transmission over a telephone line, and back again. A dial-up connection between two computers requires a modem at each end, both of which support the same communication protocols. Modems take the form of internal devices that plug into one of a computer’s expansion slots, or external devices that connect to one of the computer’s serial ports. The term “modem” is also used incorrectly, in many cases, to describe any device that provides a connection to a wide area communications service, such as a cable television (CATV) or Digital Subscriber Line (DSL) connection. These devices are not actually modems, because the service is digital, and no analog/digital conversion takes place.
MSAU See multistation access unit.
MTU See maximum transmission unit (MTU).
multicast A network transmission with a destination address that represents a group of computers on the network. Transmission Control Protocol/ Internet Protocol (TCP/IP) multicast addresses are defined by the Internet Assigned Numbers Authority (IANA) and represent groups of computers with similar functions, such as all of the routers on a network. See also broadcast, unicast.
multifunction cable tester An electronic device that automatically tests a variety of network cable properties, compares the results to established standards, and specifies whether the cable is functioning within the defined parameters for those properties.
multihomed A computer with two or more network interfaces, whether they take the form of network interface adapters, dial-up connections using modems, or other technologies. On a Transmission Control Protocol/ Internet Protocol (TCP/IP) network, each of the network interfaces in a multihomed computer must have its own Internet Protocol (IP) address.
multimode fiber A type of fiber optic cable typically used on local area networks (LANs) and supported by a number of data-link layer protocols, including standard Ethernet, Fast Ethernet, Gigabit Ethernet, and Fiber Distributed Data Interface (FDDI). Multimode fiber optic uses a light-emitting diode (LED) as a light source, unlike singlemode fiber optic, which uses a laser. Multimode fiber has a smaller bend radius, enabling it to bend around corners more easily than singlemode. As a result, multimode is better suited for relatively short distance connections than is singlemode. However, even multimode fiber can span much longer distances than most copper-based cables. See also singlemode fiber.
multiple master replication A technique usually associated with a directory service, in which identical copies of a database are maintained on various computers scattered throughout a network. In multiple master replication, users can make changes to any copy of the database, and the changes to that copy are replicated to all of the other copies. This is a complex technique, because it is possible for different users to make changes to the same record on different masters. The system must therefore have a mechanism for reconciling data conflicts in the various masters, such as using time stamps or version numbers to assign priorities to data modifications. Microsoft’s Active Directory directory service uses multiple master replication. See also single master replication.
multiplexing Any one of several techniques used to transmit multiple signals over a single cable or other network medium simultaneously. Multiplexing works by separating the available bandwidth of the network medium into separate bands, by frequency, wavelength, time, or other criteria, and transmitting a different signal in each band. Local area network (LAN) media carry only one signal, and therefore do not use multiplexing, but some networks, such as cable television (CATV) and telephone networks, do.
multiport repeater Another name for an Ethernet hub. A repeater is a physical layer device that amplifies incoming signals and retransmits them, enabling network segments to span longer distances without suffering from the effects of attenuation. A multiport repeater is a device that accepts multiple network connections. Signals arriving through any of the device’s ports are amplified and retransmitted out through all of the other ports simultaneously. All of the hubs used on Ethernet networks are multiport repeaters.
multistation access unit (MAU or MSAU) The hub used on a Token Ring network. Token Ring hubs are more complicated than Ethernet hubs because instead of repeating incoming signals out through all ports simultaneously, a MAU sends incoming signals out through each port in turn and waits for the signal to be returned by the connected computer. This forms the logical ring from which Token Ring networks get their name. To prevent breaks in the network, MAUs also perform an initialization process to insert each active computer into the ring.
multitasking The technique by which a computer with one processor executes multiple tasks simultaneously. By splitting the software processing into separate processes called threads, the processor in the computer can switch rapidly from one thread to another, devoting some of its clock cycles to each. There are two types of multitasking: cooperative and preemptive. In cooperative multitasking, the operating system passes control of the processor to each application in turn, and it is up to the application to return control to the operating system. A badly written application can fail to return control, causing the entire system to run inefficiently, or even crash. In preemptive multitasking, the operating system has complete control over the allocation of processor time to each application. Even if an application crashes, the rest of the processes continue to run normally.
name resolution The process of converting a computer or other device’s name into an address. Computers communicate using numeric addresses, but humans work better with names. To be able to send data to a particular destination identified by name in the user interface, the computer must first resolve that name into an address. On Transmission Control Protocol/Internet Protocol (TCP/IP) networks, for example, Domain Name System (DNS) names and Network Basic Input/Output System (NetBIOS) names must be resolved into Internet Protocol (IP) addresses. There are several name resolution methods that computers can use, depending on the type of name and type of address involved, including table lookups using text files such as Hosts and Lmhosts; independent processes, such as broadcast message generation; and network services, such as DNS and the Windows Internet Naming Service (WINS). See also Address Resolution Protocol (ARP).
NAT See Network Address Translation (NAT).
Nbtstat.exe A Microsoft Windows command-line utility that displays information about the NetBIOS Over TCP/IP (NetBT) connections that the system uses when communicating with other computers running Windows on a Transmission Control Protocol/ Internet Protocol (TCP/IP) network.
NCP See NetWare Core Protocol (NCP).
NDIS See Network Driver Interface Specification (NDIS).
NDS See Novell Directory Services (NDS).
NetBEUI See NetBIOS Extended User Interface (NetBEUI).
NetBIOS See Network Basic Input/Output System (NetBIOS)
NetBIOS Extended User Interface (NetBEUI) A transport protocol sometimes used by the Microsoft Windows operating systems for local area networking. NetBEUI was the default protocol in the first version of Microsoft Windows NT and in Microsoft Windows for Workgroups; it has since been replaced as the default Windows protocol by Transmission Control Protocol/Internet Protocol (TCP/IP). NetBEUI is a simplified networking protocol that requires no configuration and is self-adjusting. However, the protocol is suitable only for small networks, because it is not routable. NetBEUI identifies computers by the Network Basic Input/Output System (NetBIOS) names (or computer names) assigned during the Windows installation. Because NetBIOS uses no network identifier, there is no way for the protocol to route traffic to systems on another network.
netstat A command-line utility supplied with UNIX and Microsoft Windows operating systems. The netstat utility displays information about a Transmission Control Protocol/ Internet Protocol (TCP/IP) computer’s current network connections and about the traffic generated by the various TCP/IP protocols.
NetWare Core Protocol (NCP) A protocol in Novell NetWare’s Internetwork Packet Exchange (IPX) protocol suite that is responsible for all of the file-sharing traffic generated by Novell NetWare clients and servers.
NetWare Link Services Protocol (NLSP) A dynamic routing protocol created by Novell for its NetWare operating system. NLSP enables NetWare routers to exchange routing information with less overhead than protocols that rely on repeated broadcast transmissions, such as the Routing Information Protocol (RIP).
network access point A hardware device used on wireless local area networks (LANs) employing the infrastructure topology to provide an interface between a cabled network and wireless devices. The access point is connected to a standard network using a cable and also has a transceiver enabling it to communicate with wireless computers and other devices. See also infrastructure topology.
Network Address Translation (NAT) A firewall technique that enables Transmission Control Protocol/Internet Protocol (TCP/IP) client computers using unregistered Internet Protocol (IP) addresses to access the Internet. Client computers send their Internet service requests to a Network Address Translation- (NAT-) equipped router, which substitutes its own registered IP address for the client’s unregistered address, and forwards the request on to the specified server. The server sends its reply to the NAT router, which then relays it back to the original client. This renders the unregistered clients invisible to the Internet, preventing direct access to them. See also firewall.
network attached storage (NAS) A network data storage technology that uses a dedicated hardware device with a drive array and an embedded operating system.
Network Basic Input/Output System (NetBIOS) An application programming interface (API) that provides computers with a namespace and other local area networking functions.
Network Driver Interface Specification (NDIS) A multiprotocol device driver interface used by the Windows operating system for its network interface adapter drivers. The NDIS driver enables a single adapter and its data-link layer protocol to support traffic generated by the Transmission Control Protocol/Internet Protocol (TCP/IP), Internetwork Packet Exchange (IPX), and NetBIOS Extended User Interface (NetBEUI) protocols, in any combination.
Network File System (NFS) A standardized file sharing application used primarily by UNIX and Linux operating systems that enables one computer to mount the drives of another computer on the network into its own file system. File sharing interoperability with UNIX and Linux computers is frequently implemented in the form of an NFS product for another operating system, such as Microsoft Services for UNIX.
Network Information System (NIS) A directory service designed for UNIX and Linux computers that stores frequently-replicated configuration files in a central location, where they can be accessed by any system that needs them.
network interface adapter A hardware device that provides a computer with access to a local area network (LAN). Network interface adapters can be integrated into a computer’s motherboard or take the form of an expansion card, in which case they are called network interface cards (NICs). The adapter, along with its driver, implements the data-link layer protocol on the computer. The adapter has one or more connectors for network cables, or some other interface to the network medium. The network interface adapter and its driver are responsible for functions such as the encapsulation of network layer protocol data into data-link layer protocol frames, the encoding and decoding of data into the signals used by the network medium, and the implementation of the protocol’s Media Access Control (MAC) mechanism. network layer The third layer from the bottom of the Open Systems Interconnection (OSI) reference model. Protocols operating at the network layer are responsible for packaging transport layer data into datagrams, addressing them to its final destination, routing them across the internetwork, and fragmenting the datagrams as needed. The Internet Protocol (IP) is the most common protocol operating at the network layer, although Novell NetWare networks formerly used a proprietary network layer protocol called Internetwork Packet Exchange (IPX).
Network News Transfer Protocol (NNTP) A Transmission Control Protocol/Internet Protocol (TCP/IP) protocol used to post, distribute, and retrieve Usenet messages to and from news servers throughout the Internet.
Network Terminator 1 (NT-1) Short for network terminator, the hardware device on the client side of an Integrated Services Digital Network (ISDN) installation that provides the straight tip (S/T) interface used to connect equipment to the service, such as ISDN telephones, fax machines, and the terminal adapter that connects to a computer. In some cases, the NT-1 is a separate piece of equipment, but it can also be integrated into a single unit along with a terminal adapter for installations in which only a single computer is to be connected to the service.
Network Time Protocol (NTP) An application layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol used to synchronize the clocks in network computers.
NIC See network interface adapter.
NIS See Network Information System (NIS).
NLP See normal link pulse (NLP).
NLSP See NetWare Link Services Protocol (NLSP).
NNTP See Network News Transfer Protocol (NNTP).
node Any uniquely addressable device on a network, such as a computer, router, or printer.
nominal velocity of propagation (NVP) The speed at which signals travel through a particular length of cable. Cable testing devices such as time domain reflectometers (TDRs) use the NVP to compute the length of a particular cable segment by dividing it into the measured time needed for a generated test signal to travel to the other end of the cable and back. The manufacturer supplies the NVP for a particular cable.
normal link pulse (NLP) The signal generated by standard Ethernet network interface adapters and hubs, which the devices use to signal that they have been cabled together properly. When an adapter or hub receives the NLP signal from the device to which it is connected, it lights up a light-emitting diode (LED), which indicates that communication is taking place. See also fast link pulse (FLP).
Novell Directory Services (NDS) Formerly known as NetWare Directory Services, now known as eDirectory, the first hierarchical, object-oriented directory service to achieve commercial success. NDS was first released as part of NetWare 4.0 in 1993 and has matured into a robust product that now supports other platforms in addition to NetWare, such as UNIX and Microsoft Windows. NDS provides networks with single logon capabilities and the ability to support third-party applications through the use of schema extensions. See also directory service, schema.
NT-1 See Network Terminator 1 (NT-1)
NTFS Short for NT file system; one of the file systems included with the Microsoft Windows 2000 and Microsoft Windows NT operating systems. NTFS supports larger volumes than the file allocation table (FAT) file system supported by Microsoft Windows, includes transaction logs to aid in recovery from disk failures, and enables network administrators to control access to specific directories and files. The main drawback to NTFS is that the drives are not accessible by any operating systems other than Windows 2000 and Windows NT. If you boot the computer with an MS-DOS disk, for example, the NTFS drives are invisible.
NVP See nominal velocity of propagation (NVP).
open circuit A type of cable fault in which one or more wires is not properly connected to the proper contact at the other end of the connection. Cable testing equipment typically detects open circuits by transmitting a test signal from one end of the cable and then failing to detect it at the other end. See also short circuit.
Open Shortest Path First (OSPF) A dynamic routing protocol that exchanges information with other routers on the network to update the system’s routing table with current information about the configuration of the internetwork. OSPF is a link state protocol that evaluates routes based on their actual performance, rather than using a less accurate measurement like the number of hops needed to reach a particular destination. See also distance vector protocols, Routing Information Protocol (RIP).
Open Systems Interconnection (OSI) reference model A theoretical model defined in documents published by the International Organization for Standardization (ISO) and the Telecommunication Standards Section of the International Telecommunications Union (ITU-T). The OSI model is used for reference and teaching purposes and divides the computer networking functions into seven layers: application, presentation, session, transport, network, data-link, and physical (from top to bottom). However, the layers do not correspond exactly to any of the currently used networking protocol stacks.
operating system The primary program running on a computer, which processes input and output, runs other programs, and provides access to the computer’s hardware.
organizationally unique identifier (OUI) The three-byte hexadecimal value assigned by the Institute of Electrical and Electronic Engineers (IEEE) identifying the manufacturer of a network interface adapter. The OUI is used as the first three bytes of the adapter’s hardware address.
OSI See Open Systems Interconnection (OSI) reference model.
OSPF See Open Shortest Path First (OSPF).
OUI See organizationally unique identifier (OUI).
packet The largest unit of data that can be transmitted over a data network at any one time. Messages generated by applications are split into pieces and packaged into individual packets for transmission over the network. Each packet is transmitted separately and can take a different route to the destination. When all of the packets arrive at the destination, the receiving computer reassembles them into the original message. This is the basic functionality of a packet-switching network.
packet filter Packet filtering is a firewall technique in which a router is configured to prevent certain packets from entering a network. Packet filters can be created based on hardware addresses, Internet Protocol (IP) addresses, port numbers, or other criteria. For example, you can configure a router to allow only certain computers to access the network from the Internet, or allow your network users access to Internet e-mail but deny them access to Internet Web servers. Although typically used to prevent intrusion into a private network from the Internet, packet filtering can also be used to limit access to one of the local area networks (LANs) on a private internetwork.
packet switching A type of network communications in which messages are broken up into discrete units and transmitted to the destination. These units, called packets, can take different routes to the destination and might arrive there in a different order than that in which they were sent, but the receiving system can reassemble them in the proper order. Packet switching is what makes it possible for the computers on a local area network (LAN) to share a single network medium. If the computers transmitted entire messages at once, they could monopolize the network for long periods of time, preventing other computers from transmitting.
PAM-5 See Pulse Amplitude Modulation-5 (PAM-5).
PBX See private branch exchange.
PC Card A peripheral device standard designed for laptops and other portable computers, which enables manufacturers to create network interface cards (NICs), modems, and other devices packaged in a form approximately the size of a credit card.
PDU See protocol data unit (PDU).
peer-to-peer networking A networking system in which each computer can function both as a client and a server. Each computer also maintains its own security settings, which enables it to control access to its own resources. Peer-to-peer networking is useful on small networks, because no centralized administration is needed and users can easily maintain their own security settings. On larger networks, peer-to-peer networking is inefficient because users need a separate account for every computer they want to access, and because the access control capabilities are usually less flexible and less robust than those of a centrally administered client/server network.
phantom collision A phenomenon that occurs when excessive crosstalk on a twisted-pair cable causes a computer to detect signals on both the transmit and receive wire pairs at the same time. To the network interface adapter, these simultaneous signals indicate the existence of a packet collision, and the adapter takes the appropriate steps to clear the network of data and retransmit the supposedly damaged packet. In fact, no real collision has occurred, but the end result is the same as if one had.
physical layer The bottom layer of the Open Systems Interconnection (OSI) reference model. The physical layer defines the nature of the network medium itself, how it should be installed, and what types of signals it should carry. In the case of local area networking, the physical layer is closely related to the data-link layer immediately above it, because the data-link layer protocol includes the physical layer specifications.
Ping A Transmission Control Protocol/Internet Protocol (TCP/IP) command-line utility used to test whether a computer can communicate with another computer on the network. Ping generates Internet Control Message Protocol (ICMP) Echo Request messages and transmits them to the computer specified on the command line. The target computer, on receiving the messages, transmits them back to the sender as ICMP Echo Replies. The system running Ping then displays the elapsed times between the transmission of the requests and the receipt of the replies. Virtually every TCP/IP client implementation includes a version of Ping.
Plain Old Telephone Service (POTS) A common phrase referring to the Public Switched Telephone Network (PSTN), the standard copper-cable telephone network used for analog voice communications around the world.
Point-to-Point Protocol (PPP) A data-link layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol used for wide area network (WAN) connections, especially dial-up connections to the Internet and other service providers. Unlike its progenitor, the Serial Line Internet Protocol (SLIP), PPP includes support for multiple network layer protocols, link quality monitoring protocols, and authentication protocols. PPP is used for connections between two computers only and therefore does not need many of the features found in local area network (LAN) protocols, such as address fields for each packet and a Media Access Control (MAC) mechanism.
Point-to-Point Protocol over Ethernet (PPPoE) A Transmission Control Protocol/Internet Protocol (TCP/IP) standard that defines a method for establishing individually negotiated Point-to-Point Protocol (PPP) connections between computers on an Ethernet network and services on other networks, accessible through a Digital Subscriber Line (DSL) or cable television (CATV) connection. See also Point-to-Point Protocol (PPP).
Point-to-Point Tunneling Protocol (PPTP) A data-link layer protocol used to provide secured communications for virtual private network (VPN) connections. VPNs are private network connections that use the Internet as a network medium. To secure the data as it is transmitted across the Internet, the computers use a process called tunneling, in which the entire data-link layer frame generated by an application process is encapsulated within an Internet Protocol (IP) datagram. This arrangement violates the rules of the Open Systems Interconnection (OSI) reference model, but it enables the entire Point-to-Point Protocol (PPP) frame generated by the user application to be encrypted inside an IP datagram.
POP3 See Post Office Protocol 3 (POP3).
port A code number identifying a process running on a Transmission Control Protocol/Internet Protocol (TCP/IP) computer. Transport layer protocols, such as the Transmission Control Protocol (TCP) and the User Datagram Protocol (UDP), specify the port number of the source and destination application processes in the header of each message they create. The combination of an Internet Protocol (IP) address and a port number (which is called a socket) identify a specific application on a specific computer on a specific network. Port numbers lower than 1024 are called well-known port numbers, which are assigned by the Internet Assigned Numbers Authority (IANA) to common applications. The TCP port number 80, for example, is the well-known port number for Web servers. Port numbers 1024 and above are ephemeral port numbers, which are selected at random by clients for each transaction they initiate with a server. Alternatively, a port is a hardware connector in a computer or other network device that is used to attach cables that run to other devices.
Post Office Protocol 3 (POP3) An application layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol used by e-mail clients to download messages from an e-mail server. E-mail traffic between servers and outgoing e-mail traffic from clients to servers uses the Simple Mail Transfer Protocol (SMTP). See also Internet Mail Access Protocol (IMAP).
POTS See Plain Old Telephone Service (POTS).
PPP See Point-to-Point Protocol (PPP).
PPTP See Point-to-Point Tunneling Protocol (PPTP).
presentation layer The second layer from the top of the Open Systems Interconnection (OSI) reference model. The presentation layer is responsible for translating the syntaxes used by different types of computers on a network. A computer translates the data generated by its applications from its own abstract syntax to a common transport syntax suitable for transmission over the network. When the data arrives at its destination, the presentation layer on the receiving system translates the transfer syntax into the computer’s own native abstract syntax.
PRI See Primary Rate Interface (PRI).
Primary Rate Interface (PRI) An Integrated Services Digital Network (ISDN) service that consists of 23 64-Kbps B channels plus one 64-Kbps D channel, providing an aggregate bandwidth equal to that of a T-1 line. The B channels can be combined into a single data pipe, used individually, or in any combination. The PRI service is rarely used in the United States but is a popular business service in Europe and Japan. See also B channel, D channel, Integrated Services Digital Network (ISDN).
Private branch exchange (PBX) A private telephone network used within an organization that shares a number of outside telephone lines among its users.
promiscuous mode An operational mode available in some network interface adapters that causes the adapter to read and process all of the packets transmitted over the local area network (LAN), not just the packets addressed to it. Protocol analyzers use promiscuous mode to capture comprehensive samples of network traffic for later analysis.
protocol A documented format for the transmission of data between two networked devices. A protocol is essentially a language that a computer uses to communicate, and the other computer to which it is connected must use the same language for communication to take place. In most cases, network communication protocols are defined by open standards created by bipartisan committees. However, there are still a few proprietary protocols in use. Computers use many different protocols to communicate, which has given rise to the Open Systems Interconnection (OSI) reference model, which defines the layers at which different protocols operate.
Protocol An American Standard Code for Information Interchange (ASCII) text file found on Transmission Control Protocol/Internet Protocol (TCP/IP) systems that lists the codes used in the Protocol field of the Internet Protocol (IP) header. This field identifies the transport layer protocol that generated the data carried within the datagram, ensuring that the data reaches the appropriate process on the receiving computer. The protocol numbers are registered by the Internet Assigned Numbers Authority (IANA) and are derived from the “Assigned Numbers” Request for Comments (RFC) document.
protocol data unit (PDU) A generic term for the data constructions created by the protocols operating at the various layers of the Open Systems Interconnection (OSI) reference model. For example, the PDU created by data-link layer protocols are called frames, and network layer PDUs are called datagrams.
protocol stack The multilayered arrangement of communications protocols that provides a data path ranging from the user application to the network medium. Although based on the Open Systems Interconnection (OSI) reference model, not every layer in the model is represented by a separate protocol. On a computer connected to a local area network (LAN), for example, the protocol stack generally consists of protocols at the application, transport, network, and data-link layers, the latter of which includes a physical layer specification.
proxy server An application layer firewall technique that enables Transmission Control Protocol/ Internet Protocol (TCP/IP) client systems to access Internet resources without being susceptible to intrusion from outside the network. A proxy server is an application that runs on a computer with a registered Internet Protocol (IP) address, whereas the clients use unregistered IP addresses, causing them to remain invisible from the Internet. Client applications are configured to send their Internet service requests to the proxy server instead of directly to the Internet, and the proxy server relays the requests to the appropriate Internet server, using its own registered address. On receiving a response from the Internet server, the proxy server relays it back to the original client. Proxy servers are designed for specific applications, and the client must be configured with the address of the proxy server. Administrators can also configure the proxy server to cache Internet information for later use and to restrict access to particular Internet sites. See also firewall, Network Address Translation (NAT).
PSTN See Public Switched Telephone Network (PSTN).
Public Switched Telephone Network (PSTN) The standard copper-cable telephone network used for analog voice communications around the world. Also known as Plain Old Telephone Service (POTS).
Pulse Amplitude Modulation-5 (PAM-5) A signaling scheme used in the 1000Base-T Gigabit Ethernet variant. PAM-5 is one of the elements that makes it possible for 1000Base-T to run using standard Category 5 (CAT5) unshielded twisted pair (UTP) cable.
QIC See quarter-inch cartridge (QIC).
quarter-inch cartridge (QIC) A datastorage medium that uses cartridges containing quarter-inch-wide magnetic tape, most commonly used for system backups.
RARP See Reverse Address Resolution Protocol (RARP).
redirector A network client component that determines whether a resource requested by an application is located on the network or on the local system and sends the request either to the local I/O system or to the networking protocol stack. A computer can have multiple redirectors to support different networks, such as a Microsoft Windows network and a Novell NetWare network.
remote bridge A device operating at the data-link layer of the Open Systems Interconnection (OSI) reference model. A remote bridge is used to connect two local area networks (LANs) at different locations with a wide area network (WAN) link, such as a dial-up modem connection or a leased telephone line. By bridging the two network segments, the amount of traffic passing over the WAN is limited, which compensates for its relative slow speed and high cost. See also bridge.
Remote Monitoring (RMON) protocol A network management protocol that enables hardware and software devices to transmit status information to a central network management console.
repeater A physical layer device that amplifies network signals, enabling them to travel longer distances without suffering from the effects of attenuation. Repeaters for Ethernet networks using coaxial cable have two ports, one for incoming traffic and one for outgoing traffic. However, most of the repeaters used today have multiple ports to support networks using a star topology. The hubs used for unshielded twisted pair (UTP) Ethernet networks today are all multiport repeaters, which amplify signals as they transmit them out through all of the device’s ports simultaneously. See also attenuation, hub, multiport repeater.
Request for Comments (RFC) A document published by the Internet Engineering Task Force (IETF) that contains information about a topic related to the Internet or to the Transmission Control Protocol/ Internet Protocol (TCP/IP) suite. For example, all of the TCP/IP protocols have been documented and published as RFCs and eventually might be ratified as Internet standards. Some RFCs are only informational or historical, however, and are not submitted for ratification as standards. After they are published and assigned numbers, RFCs are never changed. If a new version of an RFC document is published, it is assigned a new number and cross-indexed to indicate that it renders the old version obsolete.
resolver Another name for the Domain Name System (DNS) client found on every Transmission Control Protocol/Internet Protocol (TCP/IP) computer. Whenever the computer attempts to access a TCP/IP system using a DNS name, the resolver generates a DNS Request message and sends it to the DNS server specified in the computer’s TCP/IP client configuration. The DNS server then takes the necessary steps to resolve the requested name into an Internet Protocol (IP) address and returns the address to the resolver in the client computer. The resolver can then give the IP address to the TCP/IP client, which uses it to transmit a message to the desired destination. See also Domain Name System (DNS).
resource record The unit in which a Domain Name System (DNS) server stores information about a particular computer. The information stored in a resource record depends on the type of record it is, but typically a resource record includes the host name of a computer and its equivalent Internet Protocol (IP) address. In most cases, administrators must manually create the resource records on a DNS server, but recent additions to the DNS standards define a method for dynamically updating the information in resource records as needed. This capability is central to the DNS functionality required by the Active Directory directory service. See also Domain Name System (DNS).
Reverse Address Resolution Protocol (RARP) An alternative mode of the Address Resolution Protocol (ARP) that enables a computer to retrieve an Internet Protocol (IP) address from an RARP server by broadcasting its hardware address. Designed for use on diskless workstations, RARP is limited in that it can receive only an IP address from the server, and not other Transmission Control Protocol/Internet Protocol (TCP/IP) configuration parameters, and also in that an administrator must manually configure the RARP server with a specific IP address for every RARP client. RARP is the progenitor of the Bootstrap Protocol (BOOTP) and the Dynamic Host Configuration Protocol (DHCP).
reverse name resolution The process of resolving an Internet Protocol (IP) address into a Domain Name System (DNS) name, which is the opposite of the normal name-to-address resolution performed by DNS servers. Reverse DNS name resolution is accomplished using an extension to the DNS namespace consisting of a domain called in-addr.arpa, which contains four levels of subdomains named using the numbers 0 through 255. These subdomains contain resource records called pointers; each pointer contains an IP address and its equivalent DNS name. A DNS server looks up an IP address by locating the domain name equivalent to the address. For example, the IP address 192.168.1.15 becomes the domain name 15.1.168.192.in-addr-arpa.
RFC See Request for Comments (RFC).
RG-8 A type of coaxial cable, also known as thick Ethernet, which is specified by the original DIX Ethernet specification as well as the later Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard. RG-8 cable is 0.405 inches thick and relatively inflexible, and is installed using a bus topology. See also coaxial cable, thick Ethernet.
RG-58 A type of coaxial cable, also known as thin Ethernet, which is specified by the original DIX Ethernet specification as well as the later Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard. RG-58 cable is 0.195 inches thick and relatively flexible, uses bayonet-Neill-Concelman (BNC) connectors to join the ends, and is installed using a bus topology. See also coaxial cable, thin Ethernet.
ring A network cabling topology in which each device is connected to the next device, forming a loop with no ends. In most cases, the ring is implemented logically by the internal wiring of a hub, and the physical network takes the form of a star. See also star, topology.
RIP See Routing Information Protocol (RIP).
RJ-11 Short for Registered Jack-11, a four-pin or six-pin modular connector that is used in telephone networking. See also RJ-45.
RJ-45 Short for Registered Jack-45, an eight-pin modular connector that is used in telephone and data networking. The majority of local area networks (LANs) today use RJ-45 connectors with unshielded twisted pair (UTP) cables. See also RJ-11.
RMON See Remote Monitoring protocol (RMON).
root name server One of a handful of servers that represent the top of the Domain Name System (DNS) namespace by supplying other DNS servers with the Internet Protocol (IP) addresses of the authoritative servers for all of the top-level domains in the DNS. When resolving a DNS name into an IP address, a DNS server that cannot resolve the name itself sends a DNS Request to one of the root name servers identified in the server’s configuration. The root name server reads the toplevel domain (that is, the last word, such as com in www.microsoft.com) from the requested name and supplies the requesting server with the IP address for that top-level domain. The requesting server then transmits the same request to the top-level domain server that the root name server supplied. The root name servers are also the authoritative servers for some of the top-level domains, so they can eliminate a step from the process and supply the address of the second-level domain’s authoritative server. See also Domain Name System (DNS), authoritative server.
routed A UNIX daemon, pronounced “route-dee,” that was the original implementation of the Routing Information Protocol (RIP), the most popular of the distance vector routing protocols. See also distance vector protocol, dynamic routing.
router A network layer hardware or software device that connects two networks together and relays traffic between them as needed. A router uses a table containing information about the other routers on the network to examine the destination address of each packet it receives, select the most efficient route to that destination, and forward the packet to the router or computer that is the next step in its path. Routers can connect two local area networks (LANs) together or provide access to remote resources by connecting a LAN to a distant network using a wide area network (WAN) link. One of the most common scenarios involves using routers to connect a LAN to the network of an Internet service provider (ISP), thus providing Internet access to all of the LAN’s users.
Routing Information Protocol (RIP) A dynamic routing protocol that enables Internet Protocol (IP) and Internetwork Packet Exchange (IPX) routers to receive information about the other routers on the network, which enables them to keep their routing tables updated with the latest information. RIP works by generating broadcast messages at frequent intervals, which contain the contents of the router’s routing table. Other routers use this information to update their own tables, thus spreading the routing information all over the network. Routers also interpret the absence of RIP messages from a particular router as a sign that it is not functioning and then remove that router from their tables after a given interval. RIP is frequently criticized for the large amount of broadcast traffic that it generates on the network, and for the limitations of its distance vector routing method, which evaluates routes based solely on the number of hops between the source and the destination. See also distance vector protocol, dynamic routing.
routing table A list maintained in every Transmission Control Protocol/ Internet Protocol (TCP/IP) computer of network destinations and the routers and interfaces that the computer should use to transmit to them. In a computer that is not a router, the routing table contains only a few entries, the most frequently used of which is the default gateway entry. On a router, the routing table can contain many entries that are either manually added by a network administrator or automatically created by a dynamic routing protocol. When there is more than one routing table entry for a specific destination, the computer selects the best route based on a metric, which is a rating of the route’s relative efficiency.
SAN See Storage Area Network.
SC See subscriber connector (SC).
schema The structure of a database system. In a hierarchical directory service, such as Microsoft’s Active Directory directory service or NetWare’s Novell Directory Services (NDS), the schema contains object classes, which specify what objects can be created in the directory, the relationships between the object classes in the directory tree, and the attributes that make up each object class. Third-party applications can expand the schema for these directory services, enabling the creation of new object classes or the addition of new attributes to existing object classes. In Active Directory, it’s also possible to modify the schema manually using the Active Directory Schema console.
scope The pool of Internet Protocol (IP) addresses on a given subnet that a Dynamic Host Configuration Protocol (DHCP) server is configured to assign to clients when using the automatic or dynamic allocation method. See also Dynamic Host Configuration Protocol (DHCP), automatic allocation, dynamic allocation.
SCSI See Small Computer System Interface (SCSI).
Secure Hypertext Transfer Protocol (S-HTTP or HTTPS) A security protocol that provides authentication and encryption services to Web client/server transactions. See also Hypertext Transfer Protocol (HTTP).
Secure Sockets Layer (SSL) A security protocol that provides authentication and encryption services to Web client/server transactions. See also Hypertext Transfer Protocol (HTTP).
segment A section of a network that is bounded by hubs, bridges, routers, or switches. Depending on the data-link layer protocol and type of cable being used, a segment can consist of more than one length of cable. For example, a thin Ethernet network uses separate pieces of coaxial cable to connect each computer to the next one on the bus, but all of those pieces of cable together are called a segment.
Sequenced Packet Exchange (SPX) A connection-oriented, transport-layer protocol in the Novell NetWare Internetwork Packet Exchange (IPX) protocol suite.
Serial Line Internet Protocol (SLIP) A data-link layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol used for wide area network (WAN) connections, especially dial-up connections to the Internet and other service providers. Because it is used for connections between two computers only, SLIP does not need many of the features found in local area network (LAN) protocols, such as address fields for each packet and a Media Access Control (MAC) mechanism. SLIP is the simplest of protocols, consisting only of a single End Delimiter byte that is transmitted after each Internet Protocol (IP) datagram. Unlike its successor, the Point-to-Point Protocol (PPP), SLIP has no inherent security capabilities or any other additional services. For this reason, it is rarely used today.
service The Microsoft Windows term for a computer program or process that runs continuously in the background and performs tasks at predetermined intervals or in response to specific events. Called a daemon by UNIX operating systems, services typically perform server tasks, such as sharing files and printers, handling e-mail, and transmitting Web files.
service-dependent filtering A type of packet filtering used in firewalls that limits access to a network based on the port numbers specified in packets’ transport layer protocol headers. The port number identifies the application that generated the packet or that is destined to receive it. With this technique, network administrators can limit access to a network to specific applications or prevent users from accessing specific applications outside the network. See also firewall, port, packet filtering.
service pack (SP) A software update package provided by Microsoft for one of its products. A service pack contains a collection of fixes and enhancements packaged into a single self-installing archive file.
Services An American Standard Code for Information Interchange (ASCII) text file found on Transmission Control Protocol/Internet Protocol (TCP/IP) systems that lists the codes used in the Source Port and Destination Port fields of the Transmission Control Protocol (TCP) and User Datagram Protocol (UDP) headers. These fields identify the application process that generated the data carried within the packet, or for which it is destined. The port numbers are registered by the Internet Assigned Numbers Authority (IANA).
session layer The third layer from the top of the Open Systems Interconnection (OSI) reference model. There are no specific session layer protocols, but there are 22 services that the session layer performs, which are incorporated into various application layer protocols. The most important of these functions are dialog control and dialog separation. Dialog control provides two modes for communicating systems: two-way alternate (TWA) mode or two-way simultaneous (TWS) mode. Dialog separation controls the process of inserting checkpoints in the data stream to synchronize functions on the two computers.
shielded twisted pair (STP) A type of cable used for local area networking in environments where additional shielding against electromagnetic interference (EMI) is needed. The cable consists of eight copper wires twisted into four pairs, with different twist rates and foil or mesh shielding around each pair. The four pairs are then encased in an insulating sheath that provides even more protection.
short circuit A type of cable fault in which two or more of the conductors inside the cable are in contact with each other. Shorts can be caused by a faulty cable installation, in which connectors are improperly attached, or a break in the insulation surrounding the cable’s conductors, due either to mishandling or a manufacturing defect. Even the most basic cable testers can easily detect shorts.
signal quality error The technical term used in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard for a packet collision, which occurs when two computers on a shared network medium transmit data at precisely the same time. See also collision.
signaled error A transmission error that has already been detected by a protocol operating at a lower layer of the networking stack.
Simple Mail Transfer Protocol (SMTP) An application layer TCP/IP protocol used to carry e-mail messages between servers and from clients to servers. To retrieve e-mail from mail servers, clients typically use the Post Office Protocol (POP3) or the Internet Mail Access Protocol (IMAP).
Simple Network Management Protocol (SNMP) An application layer Transmission Control Protocol/ Internet Protocol (TCP/IP) protocol and query language used to transmit information about the status of network components to a central network management console. Components embedded into network hardware and software products called agents are responsible for collecting data about the activities of the products they service, storing the data in a management information base (MIB), and transmitting that data to the console at regular intervals using SNMP messages.
single master replication A technique usually associated with a directory service in which identical copies of a database are maintained on various computers scattered throughout a network. In single master replication, users can make changes on only one copy of the database (the master), and the master replicates those changes to all of the other copies. This is a relatively simple technique compared to multiple master replication, because data only travels in one direction. However, the system is limited in that users might have to connect to a master located at another site to make changes to the database.
singlemode fiber A type of fiber optic cable typically used for long-distance connections between networks, supported by a relatively small number of data-link layer protocols, such as Gigabit Ethernet. Singlemode fiber optic uses a laser as its light source, unlike multimode fiber optic, which uses a light-emitting diode (LED). Singlemode fiber has a larger bend radius than multimode fiber, which makes singlemode more difficult to bend around corners. As a result, singlemode is better suited than multimode for long-distance connections.
sliding window A technique used to implement flow control in a network communications protocol. By acknowledging the number of bytes that have been successfully transmitted and specifying the number of bytes that it is capable of receiving, a computer on the receiving end of a data connection creates a “window” that consists of the bytes the sender is authorized to transmit. As the transmission progresses, the window slides along the byte stream, and might change its size, until all data has been transmitted and received successfully.
SLIP See Serial Line Internet Protocol (SLIP).
Small Computer System Interface (SCSI) A peripheral device interface that enables you to connect internal and external devices (especially storage devices) to a computer. SCSI is the preferred interface for network servers.
SMTP See Simple Mail Transfer Protocol (SMTP).
SNMP See Simple Network Management Protocol (SNMP).
SNMP agent A software component integrated into a network hardware or software product, which is designed to gather ongoing status information about the product, store it in a management information base (MIB), and transmit it to a central network management console at regular intervals, using Simple Network Management Protocol (SNMP) messages.
socket On a Transmission Control Protocol/Internet Protocol (TCP/IP) network, the combination of an Internet Protocol (IP) address and a port number, which together identify a specific application process running on a specific computer. The Uniform Resource Locators (URLs) used in Internet client applications express a socket as the IP address followed by the port number, separated by a colon, as in 192.168.1.17:80.
Source IP Address A 32-bit field in the Internet Protocol (IP) header that contains a value used to identify the particular network interface from which a packet originated.
SP See service pack (SP).
spanning tree algorithm (STA) A protocol used by network bridges in cases where a network contains redundant bridges for fault-tolerance purposes. The presence of multiple bridges on the same network, performing the same tasks, can result in data loss when each bridge lists a computer as being part of a different network segment, or can even result in a bridge loop, in which packets are forwarded endlessly from bridge to bridge. Using the STA, the redundant bridges communicate among themselves and select one of the bridges to process packets, while the others remain idle until the active bridge fails.
split pair A type of twisted-pair cable fault in which two or more wires are connected to the wrong contacts in the same way at both ends of the cable. The cable appears to be wired correctly, because each contact in one connector is connected to the equivalent contact in the other connector, but the wires are not twisted into the appropriate pairs. If two signal-carrying wires are twisted together (instead of the normal configuration, in which each signal-carrying wire is twisted together with a ground wire), the cable generates excessive amounts of crosstalk, which can result in phantom collisions or other communication problems. Because the wiring appears to be correct, split pairs are not detectable by standard cable testing devices that transmit a signal at one end of the wire and receive it at the other end. To detect split pairs, you must measure the crosstalk produced by the cable, which requires a high-end multifunction cable tester.
ST See straight tip (ST) connector.
STA See spanning tree algorithm.
star A network cabling topology in which each device is connected to a central nexus called a hub. See also topology.
stateful packet inspection A generic term for a firewall process that examines connection information and traffic patterns, as well as packet contents, to determine whether packets should be permitted to pass through the firewall.
static routing A method for creating a Transmission Control Protocol/Internet Protocol (TCP/IP) router’s routing table, in which the table entries are manually created by a network administrator. See also dynamic routing, in which routing table entries are automatically created by specialized routing protocols that exchange information with the other routers on the network.
S/T interface On an Integrated Services Digital Network (ISDN) installation, the interface provided by a Network Terminator 1 (NT-1), to which you can connect ISDN devices (like ISDN telephones or faxes) or a terminal adapter (to which you can connect standard analog communications devices). In some cases, the NT-1 and the terminal adapter are integrated into a single unit, eliminating the need for straight tip (S/T) interface connectors.
storage area network (SAN) A dedicated local area network (LAN) that connects servers with storage devices, often using the Fibre Channel protocol, reducing the storage-related traffic on the user network.
STP See shielded twisted pair (STP).
straight-through connection A twisted-pair cable wiring scheme in which each of the eight wires is connected to the same contact in the connectors on both ends of the cable. This type of cable, by itself, does not permit communications between computers to take place, because the transmit signals generated by each computer are wired to the transmit contacts in the other computer. For communication to be possible, the transmit contacts in one computer must be wired to the receive contacts in the other computer, resulting in what is called a crossover circuit. Twisted-pair Ethernet networks rely on hubs to provide the crossover circuit, which enables all of the cables to be wired straight through. To connect two computers directly, without a hub, you must use a crossover cable, which provides the crossover circuit in the cable’s wiring. See also crossover connection, crossover cable.
straight tip (ST) A connector used with fiber optic cables.
striping A data availability technique in which data is written to clusters on multiple drives in an alternating pattern (that is, one cluster is written to one drive, then the next cluster to a different drive, and so on). The drives appear as a single volume to users, but because the computer is reading data from two or more physical drives, it is possible for the heads in one drive to be moving to the next cluster while the heads in the other drive are actually reading a cluster. This speeds up the disk read process, because one of the drives is always reading data; if only a single drive were used, it would have to stop reading after every cluster so the heads could move to their next location. The drawback of the striping method is that the failure of one drive causes the loss of the entire volume.
subnet A group of computers on a Transmission Control Protocol/Internet Protocol (TCP/IP) network that share a common network identifier. In some cases, a TCP/IP network is divided into multiple subnets by modifying the subnet mask and designating some of the host identifier bits as subnet identifier bits. This enables the administrator to divide a network address of a particular class into multiple subnets, each of which contains a group of the hosts supported by the class.
subnet mask A Transmission Control Protocol/Internet Protocol (TCP/IP) configuration parameter that specifies which bits of the Internet Protocol (IP) address identify the host and which bits identify the network on which the host resides. When the subnet mask is viewed in binary form, the bits with a value of 1 are the network identifier and the bits with a value of 0 are the host identifier.
subscriber connector (SC) A connector used with fiber optic cables.
switch A data-link layer network connection device that looks like a hub, but forwards incoming packets only to the computers for which they are destined. Switches essentially eliminate the medium sharing from Ethernet networks by providing each computer with a dedicated connection to its destination. Using switches, you can build larger network segments, because there is no contention for the network medium and no increase in collisions as the number of computers connected to the network rises. See also hub, which forwards incoming packets out through all of its ports.
Synchronous Optical Network (SONET) A physical layer standard that defines a method for building a synchronous telecommunications network based on fiber optic cables. SONET provides connections at various optical carrier (OC) levels running at different speeds, ranging from 51.84 Mbps (OC-1) to 9953.280 Mbps (OC-192).
T-1 A dedicated telephone connection, also called a leased line, running at 1.544 Mbps. A T-1 line consists of 24 64-Kbps channels, which can be used separately, in combinations, or as a single data pipe. Large companies use T-1 lines for both voice and data traffic; smaller companies can lease part of a T-1, which is called a fractional T-1 service. Although it uses the telephone network, a T-1 used for data networking does not use a dial-up connection; it is permanently connected to a specific location. See also leased line.
T-3 A dedicated telephone connection, also called a leased line, running at 44.736 Mbps. See also leased line.
TCP See Transmission Control Protocol (TCP).
TCP/IP See Transmission Control Protocol/Internet Protocol.
TDR See time domain reflectometer (TDR).
TE-1 A device designed to connect directly to the straight tip (S/T) interface provided by an Integrated Services Digital Network (ISDN) installation.
TE2 A device that cannot connect directly to the straight tip (S/T) interface provided by an Integrated Services Digital Network (ISDN) installation and requires an intervening terminal adapter.
Telecommunications Industry Association/Electronic Industries Alliance A cooperative trade association responsible for the “Commercial Building Telecommunication Cabling Standard,” also known as TIA/EIA 568, which specifies how network cables should be installed in a commercial site.
Telecommunications Network Protocol (Telnet) An application layer Transmission Control Protocol/Internet Protocol (TCP/IP) client/server protocol used to remotely control a computer at another location. A mainstay of UNIX networking, Telnet is a true remote control application. When you access another computer and run a program, it is the processor in the remote computer that executes that program. The Telnet service is command-line based, making it relatively useless on computers running Microsoft Windows, which rely on a graphical interface. However, all versions of Windows include a Telnet client. Microsoft Windows 2000 also includes a Telnet server, but compared with a UNIX Telnet implementation, there are relatively few things that you can do with it.
telepole A cable installation tool that consists of a telescoping pole with a hook on the end, used for pushing cables through ceiling and wall spaces.
Telnet See Telecommunications Network Protocol (Telnet).
terminal adapter A hardware component used to connect a TE2 device to an Integrated Systems Digital Network (ISDN) connection. The terminal adapter plugs into the straight tip (S/T) interface provided by the NT-1. In some cases, a terminal adapter and an NT-1 are integrated into a single unit, which is specifically designed for installations in which a computer will be the only device using the ISDN connection. See also Integrated Services Digital Network (ISDN), NT-1, TE2, S/T interface.
termination The connection of a resistor pack to the ends of a bus network to prevent signals reaching the end of the cable from reflecting back in the other direction. All bus networks, including thick and thin Ethernet and the Small Computer System Interface (SCSI) bus used for storage arrays in computers, must be terminated at both ends, or communications will not be reliable.
thick Ethernet Also called 10Base5, an Ethernet physical layer specification that uses RG-8 coaxial cable in a bus topology, with network segments up to 500 meters long and running at 10 Mbps. Thick Ethernet was the original Ethernet physical layer option introduced in the DIX Ethernet standard and was maintained in the Institute of Electrical and Electronics Engineers (IEEE) 802.3 standard. However, because of its difficult installation, it was quickly replaced by thin Ethernet, which has now been replaced by unshielded twisted pair (UTP) cable.
thin Ethernet Also called 10Base2, an Ethernet physical layer specification that uses RG-58 coaxial cable in a bus topology, with network segments up to 185 meters long and running at 10 Mbps. Thin Ethernet was the dominant Ethernet physical layer option for several years, but it has since been replaced by unshielded twisted pair (UTP) cable, which is easier to install and maintain and can run at faster speeds.
tif A file format commonly used to store graphic images in bitmap form.
time domain reflectometer (TDR) A cable testing device that measures the length of a cable by transmitting a test signal and measuring the time it takes for the signal to travel to the other end and back. By supplying the cable’s nominal velocity of propagation (the speed at which signals travel through the cable), the TDR can compute the length of the cable. In most cases, the time domain reflectometry function is incorporated into a multifunction cable tester, but it is sometimes a separate unit. See also nominal velocity of propagation (NVP).
token passing A Media Access Control (MAC) mechanism used on ring topology networks that uses a separate frame type called a token, which circulates around the network from computer to computer. Only the computer in possession of the token is permitted to transmit its data, which prevents computers from transmitting at the same time, causing collisions. On receipt of the token, a computer transmits a packet and either regenerates a new token immediately or waits for the packet to circulate around the network and return to its source, at which time the computer removes the packet and transmits the token frame. Unlike the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) MAC mechanism, no collisions occur on a properly functioning token passing network. Token passing is used by several different data-link layer protocols, including Token Ring and Fiber Distributed Data Interface (FDDI).
Token Ring A data-link layer protocol originally developed by IBM, used on local area networks (LANs) with a ring topology. Running at 4 Mbps or 16 Mbps, Token Ring networks use the token passing Media Access Control (MAC) mechanism. Although they use a logical ring topology, Token Ring networks are physically cabled like a star, using a hub called a multistation access unit (MAU) that transmits incoming packets out through each successive port in turn. Early Token Ring networks used a shielded twisted pair (STP) cable known as IBM Type 1, but today, most Token Ring networks use unshielded twisted pair (UTP) cable.
Token Ring media filter A hardware adapter device that enables you to connect a computer with a Type 1 Token Ring network interface adapter to an unshielded twisted pair (UTP) network.
tone generator and locator An inexpensive cable testing tool that consists of a transmitter device, which you connect to a cable or a wire, which generates a test signal, and a probe that can detect the signal when you touch it to the cable or the cable sheath. You can use a tone generator to test entire cables or individual wires, but because you must test each wire individually, this is not a practical tool for the cable installer seeking to test a large number of cable runs. A tone generator and locator is also known as a “fox and hound.”
top-level domain The highest level in the Domain Name System (DNS) namespace, and the right-most word in a DNS name. For example, in the DNS name www.microsoft.com, com is the top-level domain.
topology The method used to install network cabling and connect the network computers to the cable. The topology is determined by the data- link layer protocol and cable type you choose. The three basic network topologies are the bus, in which one computer is connected to the next in daisy-chain fashion; the star, in which all of the computers are connected to a central hub; and the ring, in which the computers are logically connected to each other with the ends joined together.
traceroute A TCP/IP command-line utility that displays the path that packets are taking to a specific destination. Traceroute uses Internet Control Message Protocol (ICMP) Echo Request and Echo Reply messages with varying Time To Live (TTL) values in the IP header. This causes packets to time out at each successive router on the way to the destination, and the error messages generated by the timeouts enable the Traceroute program to display a list of the routers forming the path to the destination.
transfer syntax A format used to encode application information for transmission over a network. The presentation layer of the Open Systems Interconnection (OSI) reference model is responsible for converting application data from its native abstract syntax to a common transfer syntax understood by both communicating systems. See also abstract syntax.
translation bridge A data-link layer network connection device that connects networks using different media (such as two different types of Ethernet) or different data-link layer protocols (such as Ethernet and Token Ring). In addition to selectively propagating packets to the other network segment, this type of bridge also strips off the data-link layer protocol header and rebuilds a new one using the other protocol. See also bridge, router, transparent bridge.
Transmission Control Protocol (TCP) A Transmission Control Protocol/Internet Protocol (TCP/IP) transport layer protocol used to transmit data generated by applications, such as entire files. TCP is a connection- oriented protocol that provides guaranteed delivery service, packet acknowledgment, flow control, and error detection. The two computers involved in the TCP transaction must exchange a specific series of messages called a three-way handshake to establish a connection before any application is transmitted. The receiving computer also transmits periodic acknowledgment messages to verify the receipt of the data packets. After the data is transmitted, the two computers also perform a connection termination procedure. These additional messages, plus the large 20-byte TCP header in every packet, greatly increase the protocol’s control overhead.
Transmission Control Protocol/Internet Protocol A set of networking protocols used on the Internet that provides communications across interconnected networks that consist of computers with diverse hardware architectures and various operating systems. TCP/IP includes standards for how computers communicate and conventions for connecting networks and routing traffic.
transparent bridging A data-link layer network connection device that connects two network segments and filters packets based on their hardware addresses, which it learns automatically, only forwarding packets that are addressed to the other network segment. A transparent bridge records the address of every packet it processes to build a list of the computers on each of the network segments it connects. This prevents the network administrator from having to manually identify the computers on each network segment. See also bridge, router, translation bridge.
transport layer The middle (fourth) layer of the Open Systems Interconnection (OSI) reference model. The transport layer contains protocols providing services that are complementary to the network layer protocol. A protocol suite typically has both connection-oriented and connectionless protocols at the transport layer, providing different types of service to suit the needs of different applications. In the Transmission Control Protocol/Internet Protocol (TCP/IP) suite, the transport layer protocols are the Transmission Control Protocol (TCP) and the User Datagram Protocol (UDP).
trap A message generated by a Simple Network Management Protocol (SNMP) agent and transmitted immediately to the network management console, indicating that an event requiring immediate attention has taken place.
Trivial File Transfer Protocol (TFTP) A connectionless, application layer Transmission Control Protocol/Internet Protocol (TCP/IP) protocol that transmits data files in User Datagram Protocol (UDP) packets with no authentication and no interactive interface.
tunneling A technique for transmitting data over a network by encapsulating it within another protocol. For example, Novell NetWare networks at one time supported Transmission Control Protocol/Internet Protocol (TCP/IP) only by encapsulating Internet Protocol (IP) datagrams within NetWare’s native Internetwork Packet Exchange (IPX) protocol. The Point-to-Point Tunneling Protocol (PPTP) also uses tunneling to carry Point-to-Point Protocol (PPP) frames inside IP datagrams.
Type 1 cable A type of shielded twisted pair (STP) cable used for longer cable runs on Token Ring networks.
Type 6 cable A type of shielded twisted pair (STP) cable used for patch cable connections on Token Ring networks.
UART See universal asynchronous receiver-transmitter (UART).
UDP See User Datagram Protocol (UDP).
U interface The connection provided by the telephone company in an Integrated Services Digital Network (ISDN) installation, to which you attach an NT-1. See also Integrated Services Digital Network (ISDN), NT-1.
unicast A network transmission addressed to a single computer only. See also broadcast, multicast.
universal asynchronous receiver-transmitter (UART) A component found in internal modems and computers’ serial ports that is responsible for handling the systems’ asynchronous serial communications. High-speed external modems should always use a serial port having a 16550 UART chip. Current-production internal modems all have integrated 16550 UARTs.
universal serial bus (USB) An external peripheral bus standard that is rapidly replacing many of the other device ports commonly used on computers.
unqualified name An incomplete Domain Name System (DNS) name that identifies only the host, not the domain in which the host resides. Some Transmission Control Protocol/ Internet Protocol (TCP/IP) clients can handle unqualified names by automatically appending to them the name of the domain in which the computer is located or by appending user-specified domain names.
unshielded twisted pair (UTP) A type of cable used for data and telephone networking that consists of eight copper wires twisted into four pairs with different twist rates, encased in a protective sheath. The twisting of the wire pairs reduces the crosstalk generated by signals traveling over the wires and minimizes their susceptibility to electromagnetic interference (EMI). UTP cables are graded by the Telecommunications Industry Association/Electronic Industries Association (TIA/EIA) using a series of categories. Most UTP cable installed today is Category 5 (CAT5), although Enhanced Category 5 (Category 5e, or CAT5e) cable is also available.
unsignaled error A transmission error that has not been detected by a protocol operating at a lower layer of the networking stack.
USB See universal serial bus (USB).
Usenet An Internet bulletin board system consisting of tens of thousands of conferences, called newsgroups, covering a wide range of technical, recreational, and informational topics. Users access Usenet conferences by using newsreader software to connect to a news server. This access is usually provided by Internet service providers (ISPs).
User Datagram Protocol (UDP) A connectionless Transmission Control Protocol/Internet Protocol (TCP/IP) transport layer protocol used for short transactions, usually consisting of a single request and reply. UDP keeps overhead low by supplying almost none of the services provided by its connection-oriented transport layer counterpart, the Transmission Control Protocol (TCP), such as packet acknowledgment and flow control. UDP does offer an error detection service, however. Because it is connectionless, UDP generates no additional handshake messages, and its header is only 8 bytes long.
UTP See unshielded twisted pair (UTP).
V.90 The current standard for 56-Kbps dial-up modem communications, ratified by the International Telecommunications Union (ITU) in 1998 to reconcile the competing X2 and K56 flex standards. Virtually all modems manufactured today support the V.90 standard.
virtual LAN (VLAN) A technique often used on switched networks to make a group of computers behave as though they are connected to the same local area network (LAN), even though they are physically connected to different network segments. Computers can remain in the same VLAN even when they are physically moved to a different segment.
virtual private network (VPN) A technique for connecting to a network at a remote location using the Internet as a network medium. A user can dial into a local Internet service provider (ISP) and connect through the Internet to a private network at a distant location, using a protocol like the Point-to-Point Tunneling Protocol (PPTP) to secure the private traffic.
virus A deliberately created, potentially damaging program or routine that infects a computer from an outside source (such as a file download or a floppy disk) and then replicates itself, enabling it to infect other computers.
VLAN See virtual LAN (VLAN).
VPN See virtual private network (VPN).
WAN See wide area network (WAN).
well-known port Transmission Control Protocol/Internet Protocol (TCP/IP) port numbers that have been permanently assigned to specific applications and services by the Internet Assigned Numbers Authority (IANA). Well-known ports make it possible for client programs to access services without having to specify a port number. For example, when you type a Uniform Resource Locator (URL) into a Web browser, the port number 80 is assumed, because this is the port associated with Web servers. See also ephemeral port.
wide area network (WAN) A network that spans a large geographical area using long-distance point-to-point connections, rather than shared network media as with a local area network (LAN). WANs can use a variety of communication technologies for their connections, such as leased telephone lines, dial-up telephone lines, and Integrated Services Digital Network (ISDN) or Digital Subscriber Line (DSL) connections. The Internet is the ultimate example of a WAN. See also local area network (LAN).
Windows Internet Name Service (WINS) A service supplied with the Microsoft Windows NT and Microsoft Windows 2000 operating systems that registers the Network Basic Input/Output System (NetBIOS) names and Internet Protocol (IP) addresses of the computers on a local area network (LAN) and resolves NetBIOS names into IP addresses for its clients as needed. WINS is the most efficient name resolution method for NetBIOS-based networks because it uses only unicast transmissions. Other methods rely on the repeated transmission of broadcast messages, which can generate large amounts of network traffic.
Winipcfg.exe A graphical utility included with Microsoft Windows 95, Microsoft Windows 98, and Microsoft Windows Me that you can use to view the Transmission Control Protocol/Internet Protocol (TCP/IP) configuration parameters for a particular computer. A command-line version of the tool (Ipconfig.exe) is included with Microsoft Windows 2000 and Microsoft Windows NT. Winipcfg.exe is most useful on computers with TCP/IP clients configured automatically by a Dynamic Host Configuration Protocol (DHCP) server, because it is the easiest way to view the assigned settings for the client system. You can also use Winipcfg.exe to release and renew DHCP-assigned TCP/IP configuration parameters.
WINS See Windows Internet Naming Service (WINS).
wire map tester A relatively inexpensive cable testing device used to detect open circuits, short circuits, and transposed wires in twisted-pair cable installations. The tester consists of two units that connect to the ends of the cable. One unit transmits test signals and the other unit detects them. The wire map tester is faster and more convenient than a tone generator and locator because it tests all eight wires in a twisted-pair cable run at the same time.
X.500 A standard published by the International Telecommunications Union (ITU) and the International Organization for Standardization (ISO) defining the structure of a global directory service. Microsoft’s Active Directory directory service and NetWare’s Novell Directory Services (NDS) are both based on the X.500 design.
zip A file format that is typically used to package multiple files into a single compressed file (called an archive) for transmission over a network.
Zip The proprietary name for a magnetic cartridge drive holding 100 MB or 250 MB.