Bit mask used to select bits from an Internet address to designate the network number and subnet portions of that address. All bits that are selected are set to 1, while unselected bits are 0. This is referred to as the subnet mask.
One-to-one mapping of an IP address to a corresponding physical address.
American National Standards Institute. A non-profit organization that defines standards for the productivity and international competitiveness of American industrial enterprises. ANSI participates in defining network protocol standards, computer languages, and device drivers.
Address Resolution Protocol. This protocol is used to dynamically bind the Internet address to the Ethernet or Token Ring physical address. In a routed network, ARP remains on one physical network. ARP is a broadcast protocol.
See proxy ARP.
Advanced Research Projects Agency. Former name of DARPA, the U.S. Federal Government agency that funded the ARPANET, and later the DARPA Internet.
Long haul network funded by ARPA and built by Bolt, Beranek, and Newman (BBN). It served as the basis for early networking research as well as a central backbone during development of the Internet. The ARPANET consists of individual packet switch nodes interconnected by leased lines.
Automatic Repeat Protocol
Set of autonomous systems.
Collection of routers and networks that fall under one administrative entity and cooperate closely to propagate network reachability information among themselves using an Interior Gateway Protocol (IGP). Separate autonomous systems communicate among themselves using Exterior Gateway Protocol (EGP).
Network technology that requires a single carrier frequency and requires all stations attached to the network to participate in every transmission. Ethernet is a baseband technology.
Computer that connects two or more networks and forwards packets between them. See also gateway.
Network technology that multiplexes multiple, independent network carriers onto a single cable. This technology carries voice, video, and data over the same cable.
Message destined for all stations. Broadcasts can be hardware broadcasts or can be implemented in the higher layers, such as the IP directed broadcast (126.255.255), which would be sent to all nodes in this particular subnet.
Consultative Committee for International Telegraph and Telephone
Classless Inter-Domain Routing protocol (described in RFC 1517) that makes a radical change in assigning internet addresses to hosts. CIDR virtually eliminates network classes by assigning blocks of Class C addresses to Internet providers and having the providers “subnet” the addresses in further units to organizations. CIDR also aggregates routing information to sharply reduce the growth in routing tables in Internet routers beyond their manageable capacity.
Model of interaction in a distributed system in which a program at one site sends a request to a program at another site and awaits a response. The requesting program is called the client; the program satisfying the request is called the server.
Path between two protocol modules that provides reliable data stream delivery. A connection extends from a TCP module on one machine to one on the other.
Service typified by no end-to-end acknowledgments or connections. The User Datagram Protocol and Internet Protocol are two connectionless services.
Gateways in the core system form a central part of the internet routing in that all groups must advertise paths to their networks to core gateways using Exterior Gateway Protocol (EGP).
Carrier Sense Multiple Access. Allows multiple stations to contend for access to a transmission medium by listening for idle time.
Carrier Sense Multiple Access with Collision Detection
Defense Advanced Research Projects Agency. Formerly called ARPA.
Defense Communications Agency, now called the Defense Information Systems Agency (DISA). The U.S. Government agency responsible for installation of Defense Data Network (ARPANET and MILNET).
Defense Data Network. Used loosely to refer to the MILNET, ARPANET, and the TCP/IP protocols the military branches use. More literally, it is MILNET and associated parts of the Internet that connect military installations.
Dynamic Host Configuration Protocol
Domain Name System
Naming hierarchy that specifies labels of sites, groups, and nodes in a structured order separated by dots.
dotted decimal notation
Format of the Internet addresses. Four eight-bit numbers separated by dots and given the decimal equivalent of each eight bits.
Exterior Gateway Protocol. Used by autonomous systems to communicate between themselves and the core gateways. One gateway is designated as an EGP gateway in each autonomous system.
Wrapping or encapsulating of one protocol within another.
End system. A host or other local node.
Best-known proposal for controlling congestion in gateways by restricting every host to an equal share of gateway bandwidth.
Fiber Distributed Data Interface
Gateways can divide messages into smaller units for more efficient transmission. These units are called fragments.
File Transfer Protocol. The Internet standard for transferring files from one machine to another. FTP uses the TELNET and TCP protocols.
Term applied to both the DEC LSI-11 machine and the gateway software that runs on it.
Program under 4.3 BSD UNIX on a gateway to allow information gathering using RIP, HELLO, or other interior gateway protocols to advertise routes to another autonomous system using the Exterior Gateway Protocol (EGP).
Special purpose network processor that routes packets between two or more networks. In the case of IP, the gateway routes datagrams among directly connected or adjacent networks.
Gateway-to-Gateway Protocol. Used by core gateways to exchange routing information.
Protocol that uses minimal delay as the measure of best path. Used by fuzzballs.
The number of gateways necessary for a datagram to travel between two points. If a hop count equals 6, then there are 6 gateways between specified points.
Internet Architecture Board
Internet Control Message Protocol. A part of IP that handles error and system level messages and sends them to the offending gateway or host.
Institute of Electrical and Electronics Engineers
Interior Gateway Protocol. A protocol used to propagate network reachability and routing information within an autonomous system.
ARPANET, MILNET, and NSFnet plus a collection of other networks that are connected through gateways and implement TCP/IP.
Known as an IP address. Consists of 32 bits divided into four eight-bit sections that are separated by dots. The address format is typically dotted decimal. Each node that has a TCP/IP network interface must have an IP address for each interface present.
Internet Protocol. Defines the datagram as a unit of information passed on the network and provides the fundamentals of connectionless service. ICMP is an integral part of IP.
Unit of information passed across the network by IP. The IP datagram includes an IP header that contains the internet source and destination addresses.
Intermediate System. Router, gateway, or bridge nodes between networks.
Value representing the relative cost of a path. Metrics include hop count, delay, bandwidth, and combinations of measurement criteria.
Formerly a part of ARPANET; separated from ARPANET in 1984 to segment the military traffic of MILNET from the research traffic of ARPANET. The two are still connected to the Internet.
Maximum Transmission Unit. The largest amount of data that can be transmitted on a given physical network. MTUs in some instances are user configurable.
A host that has two or more physical network connections. Multi-homed hosts can act as gateways.
Two separate algorithms that involve congestion control on TCP/IP networks.
Mapping of a name to a corresponding address. The Internet domain name system provides this functionality.
Network File System. Developed by Sun Microsystems, Inc., it allows a user to access file systems on other machines as if they were local. It is used primarily on UNIX and with IP. NFS can generate large volumes of traffic on a network.
Network Information Center
Open shortest path first.
Route between two end stations.
Packet Internet Groper. Program that tests host reachability by using the ICMP echo request and waiting for a reply.
Post Office Protocol
Facility used by transport protocols to distinguish multiple destinations within the same host.
Facility for one machine to answer an ARP request with its own physical address as a proxy for the host the ARP was addressed to, thereby taking responsibility for routing that message. This is usually implemented in a multi-homed host environment where there is one Internet address and two physical network connections.
Reverse Address Resolution Protocol. Used by a diskless machine at start up to find its Internet address. A broadcast is sent containing the physical address and the appropriate server responds with its Internet address.
Request for Comments. A numbered series of documents that range from Internet protocol standards to proposals to observations. RFCs are available on the Internet or from the SRI NIC.
Routing Information Protocol. Defined by RFC 1058, it is used by 4.3 BSD UNIX systems as an IGP. Also implemented in ROUTED.
Remote login. Originally developed for UNIX; allows users to remotely log in to systems and interact as if they were directly attached to the remote system. Although similar to TELNET, rlogin offers fewer capabilities and less security.
Path that network traffic takes from its source to its destination.
Pronounced “route-dee.” A program originally developed for UNIX that propagates routes on a local network using RIP.
System responsible for selecting the path or paths upon which network traffic is transmitted. The selection is generated by algorithms within the router. Also known as a gateway.
There are two basic types of routing algorithms used in routers: link state and distance vector. They use different criteria to select which path or paths the routers use for network traffic.
Remote Procedure Call
Serial Line Interface Protocol
Simple Mail Transfer Protocol. The Internet standard for transferring mail across the Internet.
Simple Network Management Protocol. One of the standard network Management protocols within the Internet (RFC 1098).
A subdivision of a network. When subnetting is used, the host portion of the Internet address is divided into a subnet number and a host number. Hosts and gateways identify which bits are used for the network and subnet number through the use of a subnet mask.
Used to identify the network and subnet bits of the Internet address from the host bits. Each bit used to represent this value is enabled as I, with the host bits remaining zero.
Transmission Control Protocol. The reliable transport protocol of the Internet suite. TCP is a connection-oriented protocol and uses IP to transmit the data across the network.
Internet standard protocol for remote terminal connection service. TELNET is an application that allows the user to connect to another machine, log in, and start a remote terminal session.
Trivial File Transfer Protocol. Provides unreliable, limited function file transfer capability. It depends on UDP for datagram delivery, and can be used as a bootstrap for diskless workstations.
Type of Service. A request field within an IP datagram that defines criteria for routing based on the characteristics such as network technology or low delay.
Trailer protocol places the header at the end of the packet (therefore a trailer) so that incoming datagrams can be received on a page boundary.
Time to Live. A field in each datagram that is given a numerical value. This value is decremented by each gateway through the datagram passes until TTL equals zero when the datagram is discarded. TTL is implemented to avoid endlessly looping packets within a network.
User Datagram Protocol. In conjunction with IP provides unreliable, connectionless datagram delivery service. UDP can address specific protocol ports as a destination within a given host.
UNIX-to-UNIX Copy Program. An application allows one UNIX system to copy files to or from another UNIX system usually over dial¬up links. It is the basis for electronic mail transfer in UNIX.