Networking Guide
Chapter 7, Configuring the Network Information Service (NIS)

How NIS works

How NIS works

NIS uses a client-server architecture to distribute information from the following files on the master server by default:

NIS can also manage /usr/lib/mail/aliases if your system uses sendmail. In addition, other frequently used files (such as telephone lists and automount maps) may be selected for NIS administration.

When NIS is initialized on the master server, information from these files is compiled into a series of databases called maps. The maps are then propagated to other servers (slave and copy-only servers) on the network, where they are translated back into ASCII files. When it is necessary to update information in these files, the ASCII files are edited on the master server and the maps are modified and repropagated using NIS commands.

In the SCO implementation of NIS, master and slave servers perform all lookups of NIS-managed information locally on servers. Copy-only servers and NIS clients use remote servers. SCO clients can request NIS maps from servers to satisfy lookup requests for password and group information. This is only supported for commands that have been built using the libsocket library. All SCO-supplied commands, such as ls(C), can read NIS password and group maps when run on a client. When a new map is received on a slave or copy-only server, it is converted to the corresponding ASCII file(s) that is then available locally to requesting processes. SCO NIS servers can also satisfy lookup requests from clients running other than SCO systems.

See also: