TCP/IP (Parallel Library) Configuration and Management Manual

Glossary
HP NonStop TCP/IP (Parallel Library) Configuration and Management Manual—522271-006
Glossary-9
GGP
GGP. See Gateway to Gateway Protocol.
Gigabit Ethernet 4-port ServerNet Adapter (G4SA). A multiport ServerNet adapter that
provides 1000 megabits/second (Mbps) data transfer rates between HP
NonStop S-series systems and Ethernet LANs. The G4SA is the only LAN adapter
supported for the I/O Adapter Module (IOAM) enclosure, and it is installed in slots 1, 2,
3, 4, and 5 of an IOAM. Although the G4SA supersedes the Ethernet 4 ServerNet
adapter (E4SA), Fast Ethernet ServerNet adapter (FESA), and the Gigabit Ethernet
ServerNet adapter (GESA), it cannot be installed in an HP NonStop S-series
enclosure.
Gigabit Ethernet ServerNet Adapter (GESA). A single-port ServerNet adapter that
provides Gigabit connectivity on a NonStop S-series server. The GESA installs directly
into an existing Ethernet port, and multiple GESAs are supported in a system
enclosure.
half-duplex mode. The communications mode in which data can be transferred in both
directions, but only in one direction at a time, and in which the direction of data flow
alternates. In the Session Layer, the data token indicates which side can send data.
header. The initial part of an SPI message. The first word of this header always contains the
value -28; the remainder of the header contains descriptive information about the SPI
message, most of which is accessible as header tokens. The tokens in an SPI
message header differ according to the type of message: the header of a message that
contains a command or response differs somewhat from the header of an event
message. An application can use SSGET or EMSGET calls to retrieve the values of
header tokens, and can use SSPUT calls to change the values of some tokens.
However, there are certain basic differences between header tokens and other tokens.
See also header token.
header token. In an SPI message, a token that provides information pertaining to the
message as a whole. Header tokens differ from other tokens in several ways: they
exist in the buffer at initialization and their values are usually set by SSINIT, they can
occur only once in a buffer, they are never enclosed in a list, they cannot be moved to
another buffer with SSMOVE, and programs cannot position to them or retrieve their
values using the NEXTCODE or NEXTTOKEN operation. Programs retrieve the values
of header tokens by passing appropriate token codes to SSGET and can change the
values of some header tokens by passing their token codes to SSPUT.
Examples of header tokens for commands are the command number, the object type,
the maximum-response token, the server-version token, the maximum-field-version
token, and the checksum token. Command and response messages contain a
specified set of header tokens; event messages, a different set with some overlap. See
also SPI message.
hierarchical routing. Routing based on a hierarchical addressing scheme. Most Internet
routing is based on a two-level hierarchy in which an Internet address is divided into a
network portion and a host portion. Gateways use only the network portion until the