The original is that when the user requests to connect to the instance, Oracle creates a Server Process whose mission is to serve the requests of that user, and this Process Server remains present as long as the session is open, and Oracle can create a Server Process for each User Process that requests to connect to the instance, and this is what is called the Dedicated Server.

Oracle has no problem creating the largest number of Server Processes to serve User Processes, so that there is a User Process for each Server Process, but of course there may be limits to the speed of the Server Processes that can operate at the same time. It can be said that whenever a user requests to connect to the instance, the listener creates a server process, but if the number of server processes increases, the listener makes a row for them. In general, this problem can be avoided by creating another listener that works on another port to distribute the work between them. But the problem is not in Oracle alone, but in the operating system on which the database works, as well as the server on which the database works. All of this gives you limits in dealing with a large number of Server Processes. The operating system, memory size, processor speed, and other factors that limit the possibility of an infinite number of Process Servers made Oracle think about the subject of Shared Server, which is, in short, creating a Shared server to serve a group of Process Users instead of creating a Process Server for each User Process, which made this model a group of Benefits:

1- Reducing the number of server processes connected to the instance.

2- Increasing the number of users connected to the database as a server process is not created for each user.

3- Reducing the number of idle server processes.

4- Reducing the amount of memory used in the work as well as other resources.