This project uses Java and then tries to simulate differnt approaches on how we can effectively handle data at the server side and implements 3 main approaches to handle this problem.
For the single threaded approach the Server tries to listen to the port for any incoming clients and when there is a client request it creates a socket connection between the server and the client.
serverSocket.setSoTimeout(10000);
The timeout function is used to simulate work being done by the server and the connection being open between the client ans server during this time.
For the multithreaded implementation, we first define the Thread for each of the client and server connections being made which allows the system to carry out multiple requests at once until the request is blocked due to I/O in which case context switching takes place. This is useful as the server can now utilize all of its memory concurrently but the problem here is that it can only make new threads to a certain limit.
These problems are further solved in threadpooling.
In this approach, we start be defining a poolSize which will be the maximum threads created by the server at that time and instead of creating and deleting the existing the threads for each process, we simply have to follow an event loop and the processes are given specific thread.
When all the threads are busy, the client is sent to a queue after which it will be allocated a thread from the pool only after another process is finished or we have to context switch because of I/O conidtions.