Communication between clients and servers can take place using a variety of mechanisms—from geographic to local networks, up to communication services—between applications at the OS level. Furthermore, a client-server architecture must be independent of the physical connection method that exists between the client and the server.

It should also be noted that it is not necessary for a client-server process to reside on physically separate systems. In fact, the server process and the client process can reside on the same computing platform.

The main objective of the client-server architecture, in the context of data management, is to allow client applications to access data managed by the servers. The server (understood in a logical sense as software) is often running on a remote system (for example, in another city or on a local network).

Therefore, client-server applications are often associated with distributed processing.