Before we proceed, you should understand the basic LightDB system architecture. Understanding how the parts of LightDB interact will make this chapter somewhat clearer.
In database jargon, LightDB uses a client/server model. A LightDB session consists of the following cooperating processes (programs):
A server process, which manages the database files, accepts
connections to the database from client applications, and
performs database actions on behalf of the clients. The
database server program is called
lightdb
.
The user's client (frontend) application that wants to perform database operations. Client applications can be very diverse in nature: a client could be a text-oriented tool, a graphical application, a web server that accesses the database to display web pages, or a specialized database maintenance tool. Some client applications are supplied with the LightDB distribution; most are developed by users.
As is typical of client/server applications, the client and the server can be on different hosts. In that case they communicate over a TCP/IP network connection. You should keep this in mind, because the files that can be accessed on a client machine might not be accessible (or might only be accessible using a different file name) on the database server machine.
The LightDB server can handle
multiple concurrent connections from clients. To achieve this it
starts (“forks”) a new process for each connection.
From that point on, the client and the new server process
communicate without intervention by the original
lightdb
process. Thus, the
master server process is always running, waiting for
client connections, whereas client and associated server processes
come and go. (All of this is of course invisible to the user. We
only mention it here for completeness.)