DIGITAL TCP/IP Services for OpenVMS
Concepts and Planning

OPCOM process: The system process that manages Operator Communication Manager (OPCOM) operations.

open network: A network made up of nonproprietary, interoperable systems.

Operator Communication Manager: A system administration tool for communicating with users and operators on the system.

OSPF (Open Shortest Path First) : One of a class of interior gateway protocols, described in more detail in the OSPF section of gated.proto(4).

open system: A nonproprietary, interoperable system with communications software.

Open System Interconnection (OSI): A suite of protocols, designed by ISO committees, to be the international standard of computer network architecture.

OpenVMS cluster: A configuration of OpenVMS processors.

OpenVMS cluster alias: An alias that allows remote hosts to address the cluster members as one host, as well as any cluster member individually.

OpenVMS file system: The OpenVMS files and directories on a mounted OpenVMS volume. These files and directories reside on a Files-11 On-Disk Structured (ODS-2) disk.

origination: The beginning point of communications on a circuit.

overmounting: The process of NFS mounting another directory over an existing mount point. The original file system is dismounted from the mount point, and the new file system is mounted.

packet: A unit of data sent across a network.

Packet Internet Groper (PING): A program used to test reachability of a destination by sending an ICMP echo request and waiting for a reply. See also loopback.

packet looping: A condition in which a packet revisits a node. See also aged packet.

packet size: The amount of data in a packet.

packet switching: A communication paradigm in which packets are individually routed between hosts, with no previously established communication path.

path: The physical lines between source nodes and destination nodes; can comprise a sequence of connected nodes. The path that the data takes through the network is transparent to users.

path cost: The sum of the circuit costs along a path between two nodes.

An OSPF (Open Shortest Path First) protocol metric. See metric and OSPF.

path length: The total distance (the number of circuits) between a source node and a destination node, measured in hops. Each line between systems, including routing nodes and end nodes, equals one hop. See also network diameter.

path name: A unique designation that identifies a directory or subdirectory. UNIX path names are composed of a series of fields separated by slashes (/); each field designates a file name that is uniquely contained in the previous field (directory).

path MTU: The smallest MTU of any data link that packets traverse between two hosts. The path MTU depends upon the route being used at the time. Therefore, the sending path MTU may differ from the receiving path MTU.

path MTU discovery: A mechanism to determine the path MTU at any one time.

path splitting: The ability to split the transmission load destined for a single node over several paths of equal path cost. Any destination node receiving data that has been split over several paths must support out-of-order packet caching.

PC-NFS Daemon: The server software that handles authentication and printing requests from personal computer implementations of NFS.

peer: Another router with which implicit or explicit communication is established by a routing protocol. Peers are usually on a shared network, but not always. This term is mostly used by BGP. Usually synonymous with neighbor.

physical address: A unique address of each physical connection of a node to the physical medium.

physical connection: The Physical layer communications path between two systems.

physical connectivity: The Physical layer connectivity that is a result of nodes being attached to each other via active lines and nodes.

PING: See Packet Internet Groper.

point-to-point circuit: A circuit that connects only two nodes. A point-to-point configuration requires a separate physical connection between each pair of nodes. Point-to-point systems communicate directly with other systems. Contrast with multipoint circuit.

point-to-point line: A line that connects two systems by using a single circuit.

Point-to-Point Protocol (PPP): A method for transmitting datagrams over serial point-to-point lines where a line is established between a remote host (usually over a telephone line) and another host acting as a gateway to a remote host.

poll: The sending of an NTP packet from a host to an NTP time server to request the current time. The server responds by recording the current time in the packet, then sending it back to the originating host. See also ``NTP packet''.

polling: Connecting to another system to check for things such as mail or news.

POP: See Post Office Protocol.

port: The end point of a communication link between two processes.

A UDP or TCP port number. Valid values are from 1 to 65535 inclusive.

port number: A 16-bit number used to identify applications using TCP or UDP. The number is stored in the Transport layer protocol headers to identify the application.

Portmapper Service: A service that client programs can use to determine the port number that another service uses. Clients use the Portmapper Service for NFC, PC-NFS, and RPC applications.

post: To send a message to a mailing list or newsgroup. Distinguished in context from "mail."

Post Office Protocol (POP): The TCP/IP-based protocol for client stations to read mail from a server.

PPP: See Point-to-Point Protocol.

PPP client: A host requiring a temporary PPP connection to a dialup provider or a terminal server.

PPP dialup provider: A host that answers modem calls from PPP clients, assigns IP addresses and establishes PPP connections initiated by PPP clients.

preference: A preference is a value from 0 to 255 used to select a route from many routes to the same destination. The route with the best (numerically lowest) preference is selected as the active route. The active route is the one installed in the kernel forwarding table and exported to other protocols. Preference zero is usually reserved for routes to directly attached interfaces. A default preference is assigned to each source from which GATED receives routes.

prefix: A contiguous mask covering the most significant bits of an address. The prefix length specifies how many bits are covered.

primary server: A BIND name server that maintains the database for a zone; secondary servers copy their information from primary servers. See also BIND server, cache server, forwarder server, and secondary server.

printcap database: The DIGITAL TCP/IP Services for OpenVMS database that maps local queues to printers on remote hosts; specifies local queues for LPD printing from remote hosts. Equivalent to the UNIX /etc/printcap file.

privileged port: A port in which the remote host has done some level of checking against the application using the port; privileged port numbers range from 1 to 1023.

process: The context within a system in which a specific computing session occurs; provides the context in which an application executes.

protocol: A set of rules that controls the communications between computers. Also, a set of conventions between communicating processes regarding the format and contents of messages to be exchanged.

Protocols can describe low-level details of machine-to-machine interfaces, such as the order in which the bits from a byte are set across a wire, or high-level exchanges between applications programs such as the way in which two programs transfer a file across the Internet.

protocol data unit (PDU): The unit of data sent across a network. Also called a packet.

protocol machine: The set of data structures and routines that implements a specific protocol and controls the progress of a communication between peer entities.

protocol overhead: The part of communications data or processing not directly consumed by the users but necessary to successfully bring about the transfer of user information.

protocol port: An abstraction that transport protocols use to distinguish among multiple destinations within a given host computer. Internet protocols identify ports using small positive integers. Usually the operating system allows an application program to specify which port it wants to use. Some ports are reserved for standard services such as electronic mail.

protocol transparency: The quality in a communications device or system that allows various higher-level protocols to coexist on the same wire. The protocols are transparent to the device or system.

protocol sequence: An ordered list of protocol identifiers.

protocol stack: The set of functions, one at each layer of the protocol stack, that work together to form a set of network services; each layer of the protocol stack uses the services of the module beneath it.

protocol transparency: The degree to which users of underlying protocols are aware of the specifics of those protocols.

proxy: The mechanism whereby one system acts on behalf of another system in responding to protocol requests. DIGITAL TCP/IP Services for OpenVMS uses a proxy mechanism to provide an OpenVMS identity (account) for each UNIX client by adding the name and identification codes of the client to a proxy database.

proxy ARP: The technique in which one machine, usually a router, answers Address Resolution Protocol (ARP) requests intended for another machine. By "faking" its identity, the router accepts responsibility for routing packets to the "real" destination. Proxy ARP allows a site to use a single IP address with two physical networks. Creating a subnet would normally be a better solution.

proxy database: The database that provides OpenVMS identities for remote NFS clients and UNIX-style identities for local NFS client users; provides proxy accounts for remote processes.

pseudodevice: A software device used to implement special-purpose transports and not directly associated with hardware.

pseudo-interface: A method of extending subnet routing using a network interface. Each network interface has one name and at most nine pseudo-interface names. Each network interface and pseudo-interface has its own IP address, network mask and broadcast mask.

public domain: Intellectual property available to users that does not require payment of a fee.

quality of service (QoS) : The OSI equivalent of TOS.

RARP: See Reverse Address Resolution Protocol.

RCD: See RMT/RCD.

RCP: See remote copy program.

reachable node: The node to which the local node has a usable communications path.

read access: The access right that grants the ability to view data.

reassembly: The process of piecing together datagram fragments to reproduce the original datagram based on the fragmentation data in the IP header of the datagram.

reassembly time: A routing parameter that can be set to specify the length of time allowed for the reassembly of a message received in fragments. If the reassembly time expires before all fragments are received, the fragments are discarded.

Record Management Services (RMS): The OpenVMS data management subsystem that defines the rules that govern the internal organization of and the methods of accessing file data, including how files are named and cataloged in directories.

reliability: The ability of a protocol to recover data that is damaged, lost, duplicated, or delivered out of order.

relative path name : A path name that does not start at the root; default directory is merged with the relative path name to form the absolute path name.

remote boot (BOOTP): The software that supports the downloading of system images and other types of files to requesting clients.

remote copy program (RCP): The program based on the Berkeley UNIX (see BSD) rcmd protocol that permits files to be copied from one computer to another by an extension to the syntax of the UNIX cp (copy) command. (RCP) does not provide the word-length adaptability and flexibility that the FTP protocol does.

remote line printing (LPR/LPD): The remote printing services for UNIX and OpenVMS client hosts.

remote node: A node in the network other than the local node.

remote file system: A file system that resides on a network host other than the local node.

remote procedure call (RPC): A programming interface for implementing the client/server model of distributed computing. In general, a request is sent to a remote system to execute a designated procedure, using arguments supplied, and the result returned to the caller. See also Sun RPC.

remote shell: A program that sends a command, shell, script, or command procedure to a remote host for execution.

remote task: A task either executing at a remote host or originating there.

repeater: A bidirectional device that amplifies or synchronizes signals into standard voltages, currents, and timing; propagates electrical signals from one Ethernet to another without making routing decisions or providing packet filtering; Physical layer intermediate system. See also bridge and router.

Request for Comments (RFC): A series of documents, begun in 1969, that describes the Internet suite of protocols and related experiments. Very few RFCs describe Internet standards, but all Internet standards are written as RFCs.

resolver: A mechanism or process to correlate a network host name into an appropriate network address in support of network applications---a network name resolver. See BIND resolver.

reserved port: An assigned port that provides services to unknown callers by providing a service contact point; reserved port numbers range from 1 to 255.

resynchronization: A process that enables the recovery of user information lost or corrupted during transfer across an association. Sets the association back to the state it was in at a specified point in the transfer.

retransmission: A method of error recovery in which stations receiving messages acknowledge the receipt of correct messages and, on receipt of incorrect messages, either do not acknowledge or acknowledge in the negative. The lack of acknowledgment or receipt of a negative acknowledgment indicates to the sending station that it should transmit the failed message again.

Reverse Address Resolution Protocol (RARP): The TCP/IP protocol that provides the reverse function of ARP. This protocol maps a physical (hardware) address to an IP address. Often used by diskless nodes when they first initialize to find their Internet address.

reverse domain: An Internet domain that BIND servers use to map IP addresses to domain names.

RFC: See Request for Comments.

RFC 822: The TCP/IP standard format for electronic mail message headers; often referred to as "822 messages". The name comes from RFC 822 that contains the specification; previously known as 733 format.

RIB (routing information base) : routing database

RIP: See Routing Information Protocol.

rlogin: Remote login: The Berkeley 4.3 BSD service that allows users of one machine to connect to other systems across the Internet and interact as if their terminals are connected the machines directly.

RMS: See Record Management Services.

RMT/RCD: Remote command that allows remote users to access magnetic tapes and CD drives.

root: The top level directory in a UNIX-style file system; also used to indicate a user (the superuser) who has special privileges. See superuser.

root mode: The file protection placed on a container file when it is created.

root name: The element of a path name that identifies the target file system.

root server: An Internet name server that knows about all of the top-level domains on the Internet network; the master servers for the Internet root zone.

round-trip delay: The total time during communications that implement a protocol with positive acknowledgments, for a message to be transmitted, arrive at its destination, and its corresponding acknowledgment to be sent and subsequently received by the sender of the original message.

The time it takes for a host to send an NTP packet to another host and get an NTP packet back from that host in reply.

round-trip time (RTT): A variable computed during TCP sessions that indicates the total time required to send a TCP segment to a remote host and receive a reply.

route: The path over the network that information takes to get from one source to its destination.

route aggregation:

route through: Data packets not destined for the local node.

routes database: The DIGITAL TCP/IP Services for OpenVMS database that specifies Internet gateways.

ROUTED: See Route Daemon.

Route Daemon (routed): A program that runs under 4.2BSD/4.3BSD UNIX systems (and derived operating systems) to propagate routes among machines on a local area network using the Routing Information Protocol; pronounced "route-d."

One of a class of interior gateway protocols, described in more detail in the RIP section of gated.proto(4).

router: A node that can send and receive data and also forward data to other nodes.

router advertisement: A Router Discovery Protocol message sent out by Router Discovery Servers to announce their existence to hosts. The router advertisement contains a list of all router addresses on a given interface and their preferences for use as a default router.

Router Discovery Protocol: An IETF standard protocol used to inform hosts of the existence of routers. It is used in place of or in addition to statically configured default routes in hosts. The protocol has a server portion that runs on routers, and a client portion that runs on hosts.

router id: A 32-bit number assigned to each router running the OSPF protocol. This number uniquely identifies the router within the autonomous system.

router_id: An IP address used as unique identifier assigned to represent a specific router. This is usually the address of an attached interface.

router solicitation: A Router Discovery Protocol message sent out by a host to request router advertisement responses from a router.

routing: A Network layer function, implemented in intermediate systems, that determines the path along which data travels to its destination and the movement of that data. See also decision.

routing database: The database that contains routing information, including destination host names, IP addresses for the hosts, gateway host names, and IP addresses for the gateways. There are two route databases: the static route database that is maintained on disk and the volatile database in memory.

The repository of all of gated's retained routing information, used to make decisions and as a source for routing information that is propagated.

routing domain: A set of hosts and routers within a single administrative domain that operates according to the same routing procedures.

Routing Information Protocol (RIP): The protocol that enables gateways to broadcast their current routing database to hosts and networks that are connected directly to them. DIGITAL TCP/IP Services for OpenVMS software implements the RIP through its dynamic routing Zserver.

One of a class of interior gateway protocols, described in more detail in the RIP section of gated.proto(4).

Routing Protocol: A protocol sent between routers by which routers exchange information on how to route to various parts of the network. The TCP/IP family of protocols has many of this type of protocol, such as RIP, EGP, BGP, OSPF, and dual IS-IS.

routing socket: A data structure used by processes to communicate routing information to the kernel. A process can add and delete routes, dump the routing table, and read messages from the kernel. The only type of socket supported in the AF_ROUTE domain is a raw socket.

routing table: The repository of all of gated's retained routing information, used to make decisions and as a source for routing information that is propagated.

RPC: See remote procedure call and Sun RPC.

rshell: Remote shell; a remote utility that gives the user with a shell session on a remote host.

RTL: See Run-Time Library.

RTT: See round-trip time.

Run-Time Library (RTL): A collection of OpenVMS procedures available to native mode images at run time; provide support routines for high-level language compilers.

SCALE: A TCP window scaling option; allows window information to be interpreted as being scaled by 1 to 16 powers of 2, thus increasing the size of the effective window.

secondary server: A master BIND server that receives authoritative database information from a primary server. See also BIND server, cache server, forwarder server, and primary server.

segment: A unit of data exchanged by the TCP modules.

segment length: The amount of sequence number space occupied by a segment, including controls that occupy sequence space.

sequence number: A 32-bit field in the TCP header that contains the sequence number of a sequenced control flag, the first byte of data, or empty segments (The sequence number of the next data octet to be sent).

serial device: A device that uses serial transmission; that is, transmits data one bit at a time on a single channel as opposed to parallel transmission, which transmits one or more bits at a time on one or more channels. Typically, terminals and printers are serial devices.

Serial Line Internet Protocol (SLIP): A protocol designed to allow a host to connect to another host over serial lines, such as telephone circuits or RS-232 cables.

server: A process that offers a service to another process over the network and accepts requests from other processes, known as clients.

service: (1) A task that an application can carry out. (2) The interface provided by a service element or layer for accessing one or more function.

service interface: The boundary at which a layer provides a service to the adjacent higher layer in the network architecture; may vary between implementations.

service parameter: The means by which a service user and a service provider exchange information.

service provider: In network architecture, the service element or layer that provides a set of services to the layer immediately above.

service specification: An international standard that describes the functions and service parameters of every service of a service provider.

service user: An application program, service element, or Network layer that uses the services of a service provider.

services database: The DIGITAL TCP/IP Services for OpenVMS database created by default that contains one entry for each service configured.

Simple Mail Transfer Protocol (SMTP): An Internet standard protocol for transferring electronic mail messages from one machine to another; specifies how two mail systems interact and the format of control messages they exchange to transfer mail.

Simple Network Management Protocol (SNMP): The network management protocol of choice for TCP/IP-based internets; allows remote monitoring and management of network devices (particularly routers and servers) from across an Internet.

simplex: An interface may be marked as simplex either by the kernel or by interface configuration. A simplex interface is an interface on a broadcast medium that is not capable of receiving packets it broadcasts.

The gated daemon takes advantage of interfaces that are capable of receiving their own broadcast packets to monitor whether an interface appears to be functioning properly.

skew: A measure, in hertz, of the difference between the actual frequency of a clock and what its frequency should be to keep perfect time. See also ``drift''.

slave server: A name server that has no access to the Internet and relies on forwarder servers to resolve queries that it cannot resolve locally. As slave servers receive information from forwarder servers, they store that information in their cache. See also cache server, forwarder server, primary server, and secondary server.

slew: To adjust gradually the time of a clock until it tells the correct time. Compare with step.

SLIP: See Serial Line Internet Protocol

SMI: See Structure of Management Information.

SMTP: See Simple Mail Transfer Protocol.

SNMP: See Simple Network Management Protocol.

socket: The end point of communication to which an IP address and port may be bound. When writing an application, it is a data structure that is part of the Internet pseudodevice created every time an OpenVMS process assigns a communication channel. The other part of the Internet pseudodevice is the device socket.