Threads vs Workers in CherryPy

Question

The idea of threads and workers is unclear to me. We can assign the no of threads associated for the CherryPy server as,

server = Server()
server.socket_port = port
server.socket_host = "0.0.0.0"
server.thread_pool = 100

There is no property for defining the workers, so does it mean that the server I am instantiating is the only worker and the threads associated in the thread_pool property is the maximum number of threads that it can utilize while serving the application (Mostly concurrent requests). Or is there a way to define the number of workers in CherryPy.

Any help is appreciated!

Answer 1

I can understand your confusion, because Cherrypy is notoriously poorly documented.

To understand the concepts of worker and thread , we need to step back and learn the structure of a typical cherrypy application.

When you run your server application, you start a process. That process contains a main thread, which is the cherrpy.engine . cherrypy.engine starts the server, which runs in its own thead. cherrypy.tree holds an hierachy of page handlers, and each of those handler may be a worker . But whether that worker has its own thread to run, or has to wait until other workers finish and free up a thread, is totally dependent on how many threads are left. Once a worker acquires a free thread , it becomes a worker thread and it immediately starts to work.