Client server model

Most web applications are written assuming a client , One is the server . The server provides specific services for clients . Can be divided into Repetitive type and And hairstyles .

Repetitive type ：

First step Wait for a client request
The second step Handle client requests
The third step Send a response to the customer who initiated the request
Step four Step one back

advantage	shortcoming
Easy to understand and learn	In the second step, you cannot provide services to other clients , You need to wait for the current client to finish processing and reach the fourth step before accepting the request of the new client

And hairstyles ：

First step Waiting for the arrival of a customer request .
The second step Start a new server to process the client's request . During this time a new process may be generated 、 Mission
Or thread , And rely on the support of the underlying operating system . How this step works depends on the operating system . New server generated
Handle all customer requests . After processing , Terminate this new server .

The third step Step one back .

advantage	shortcoming
Use the method of generating other servers to process customer requests if the operating system allows multitasking , Then you can serve multiple customers at the same time .	Need to support multitasking for the operating system

Generally speaking, for TCP The server is parallel , about UDP Server is repetitive .

reference ：
《TCP-IP Detailed explanation Volume 1 》