Attachment Unit Interface (AUI) Connector: Also called thicknet connector. The AUI is a transceiver cable that provides a path between a node's Ethernet interface and the Media Access Unit (MAU).

Bandwidth: The difference between the highest and the lowest frequencies of a transmission channel. A measure of the information capacity of the transmission channel. Bandwidth is expressed in bits per second (bps).

BNC Connector: Acronym stands for British Naval Connector. A standard connector from a thin coaxial cable to a transceiver.

Bridge: A device that interconnects local or remote networks across all higher-level protocols. Bridges form a single logical network, centralizing network administration. Bridges operate at the physical and link layers of the Open Systems Interconnect (OSI) reference model.

Cabling: The wire medium (usually Shielded Twisted Pair, Cat 5, FTP 6) by which nodes on a LAN are connected.

CAD/CAM: Computer Aided Design/Computer Aided Manufacturing.

Client/Server: A common form of distributed system in which software is split between server tasks and user or client tasks. A client sends requests to a server asking for information or action, and the server responds. There may be either one centralized server or several distributed ones.

Coxial cable: A type of electrical cable in which a solid piece of metal wire is surrounded by insulation, which in turn is surrounded by a metal tube. Coaxial cables have wide bandwidths and carry many data, voice and video signals simultaneously.

Collapsed Backbone: Network Architecture under which the backplane of a device such as a hub performs the function of a network backbone; the backplane routes traffic between desktop nodes and between other hubs serving multiple LANs.

Duplexing: Duplexing has all the benefits of mirroring but adds further protection utilizing different disk channels to connect two drives in a mirrored set. Often this is done using two SCSI controllers, one to each SCSI drive.

ECC (error checking and correction): Detects errors in transmitted or stored data and corrects them on the fly. The simplest form of error detection is a single added parity bit or a cyclic redundancy check. Multiple parity bits not only detect that an error has occurred, but also which bits have been inverted, and should therefore be re-inverted or fixed to restore the original data.

Ethernet: IEEE-standard data link protocol that specifies how data is placed on and retrieved from a common transmission medium. Data is broken into packets, which are then transmitted using the Carrier Sense Multiple Access/Collision Detect (CSMA/CD) algorithm until they arrive at the destination without colliding with any other. A node is either transmitting or receiving at any instant. Bandwidth ~10Mbit/s. Disk-Ethernet-Disk transfer rate with TCP/IP is typically 30 kilobyte per second. The cable is a 50-ohm coaxial cable with multiple shielding. Forms the underlying transport vehicle used by several upper-level protocols, including TCP/IP and XNS.

Fault tolerance: Designed into disk array subsystems to maintain data integrity and data availability before, during and after a failure. Fault tolerance implies that any component in a subsystem can fail and the subsystem will remain operational. In addition to the disks in any array subsystem, the cabling, controllers, adapters, and power supplies can have redundant capabilities.

Fiber Distributed Data Interface (FDDI): A 100 Mbit/s standard LAN architecture. The underlying medium is fiber-optic cable (though it can be copper cable, in which case it may be called CDDI) and the topology is a dual-attached, counter-rotating token rings.

FDDI rings are normally constructed in the form of a "dual ring of trees". A small number of devices, typically infrastructure devices such as routers and concentrators rather than host computers, are connected to both rings. Host computers are then connected as single-attached devices to the routers or concentrators. The whole dual ring is typically contained within a computer room.

Fiber optic cable: a transmission medium that uses glass or plastic fibers, rather than copper wire, to transport data or voice signals. The signal is imposed on the fibers via pulses (Modulation) or light from a laser or a light-emitting diode (LED). Because of its high bandwidth and lack of susceptibility to interference, fiber-optic cable is used in long-haul or noisy applications.

Gateway: Device that can interconnect networks with different, incompatible communications protocols. The gateway performs a layer-7 protocol-conversion to translate one set of protocols to another. A gateway operates at Open Systems Interconnection (OSI) layers up through the Session Layer.

Hamming: Redundant bits added to stored or transmitted data for the purposes of error detection and correction.

Hot Swap: The ability to replace failed hardware without turning off the system. The subsystem can remain operational and data is available at all times even during service activities. An example of hot swapping would be hard disks that can be added and removed while a server is operational.

Hub: A device connected to several other devices, also called a repeater. Strictly, it is a non-retiming device.

JBOD: JBOD refers to several disks in an array not set up as a RAID configuration. The acronym stands for "just a bunch of disks."

Local Area Network (LAN): A data communications network that is geographically limited allowing easy interconnection of workstations and servers within adjacent buildings. Ethernet and FDDI are examples of standard LANs. Because the network is known to cover only a small area, optimizations can be made in the network signal protocols that permit data rates up to 100Mb/s.

Metropolitan Area Network (MAN): A data network intended to serve an area the size of a large city.

Mirror: Disk mirroring provides and identical twin of all data written to the primary disk. Total usable disk space is 50% of all drives in the mirrored set. If one disk fails, the system uses the complete data from the other disk.

Network Interface Card (NIC): Also called an adapter card. A board installed in a computer to provide a physical connection to and from that computer system.

Network Operating System (NOS): The software that controls the operation of the network. A NOS enables users to communicate and to share files and peripherals. It provides the user interface to the LAN and communicates with the LAN hardware or network interface card (NIC).

Open Systems Interconnection (OSI): OSI is the umbrella name for a series of non-proprietary protocols and specifications. The OSI architecture is split between seven layers, from lowest to highest:

Physical layer: this layer determines how signals are transmitted on the network cabling.

Data Link: incorporates the logical link (LLC) and the media access control (MAC) sublayers. The data link layer transmits data groups into frames using the Ethernet or Token ring access methods.

Network layer: handles the routing of data in packets using the networking protocols.

Transport layer: ensures error free data transmissions.

Session layer: establishes and maintains connection between nodes according to the appropriate protocol.

Presentation layer: handles data encoding and formatting; provides data compression.

Application layer: provides the means for application processes to use the network services; the interface to user database, file and email software often implemented with API's (application programming interfaces).

Each layer uses the layer immediately below it and provides a service to the layer above.

Parity: Error checking procedure in which the number of 1s in a binary sequence must always be the same, either odd or even, for each group of bits transmitted. ECC adds correcting function to basic parity.

Protocol: A set of formal rules describing how to transmit data, especially across a network. Low-level protocols define the electrical and physical standards to be observed, bit- and byte-ordering and the transmission and error detection and correction of the bit stream. High-level protocols deal with the data formatting, including the syntax of messages, the terminal to computer dialogue, character sets, sequencing of messages etc.

Redundant Array of Independent (/Inexpensive) Disks (RAID): A technology using a software or hardware controller with several disk drives to allow varying degrees of either increased performance or data integrity. Levels of redundancy or data security are dependent on the number of drives in the array, as well as the way the data is stored across the drives.

Repeater: A device which propagates electrical signals from one cable to another. Less intelligent than a bridge, gateway or router.

Router: A device which forwards packets between networks. The forwarding decision is based on network layer information and routing tables, often constructed by routing protocols.

Server: A Computer that provides service for other computers connected to it via a network. The most common example is a file server that has a local disk and services requests from remote clients to read and write files on that disk using the Network File System (NFS) protocol or network operating system software.

Shielded Twisted Pair (STP):-Common transmission medium that consists of a receive (RX) and a transmit (TX) wire twisted together to reduce crosstalk. The twisted pair is shielded by a braided outer sheath.

Striping: Combines area of multiple disks into one large logical drive. Data is distributed evenly over drives in a stripe set. All drives in a stripe set work to perform the same functions done by a single drive in a normal configuration. Allows concurrent I/O. Striping requires at least 2 drives.

Thinnet: 10BASE2 standard cable. Also called cheapernet in reference to this less expensive, thinner version of traditional Ethernet cable.

Token Ring (TR): A communication methods that uses a token to control access to the LAN. The difference between a token bus and a token ring is that a token ring LAN does not use a master controller to control the token. Instead, each computer knows the address of the computer that should receive the token next. When a computer with the token has nothing to transmit, it passes the token to the next computer in line.

Topology: A network topology shows the hosts and the links between them. A network layer must stay abreast of the current network topology to be able to route packets to their final destination.

Unshielded Twisted Pair (UTP): Normal telephone wire (in the U.S.). It may be used for computer-to-computer communications. It is much less expensive than standard Ethernet cable.

Wide Area Network (WAN): Public or private computer network serving a wide geographic area.

Workstation: A general-purpose computer designed to be used by one person at a time.