Multiprocessing.Process not starting without any error

Question

Below is my demo:

Thread xxx:
    import xxx
    import xxxx

    def test_process1(hostname,proxy_param):
        # just never run
        try:                             # breakpoint 0
            with open("/xxx","w+") as f: # breakpoint 1
                f.write("something")
        except Exception as e:
            pass # just never run  breakpoint 3
    def test():
        try:
            a = Process(target=test_process1, args=(hostname,proxy_param))
            a.start()
            a.join() # you are blocking here. test_process1 not working and never quit
        except Exception as e: 
            pass # breakpoint 4

function test_process1 just never run. No error, No breakpoint.

The test function code is in a big project, here is a demo.

Answer 1

Hope. this piece of code helps.

Workers list will get divided based on the number of processes in use.

Sample Code with ManagerList.

from subprocess import PIPE, Popen
from multiprocessing import Pool,Pipe
from multiprocessing import Process, Queue, Manager

def child_process(child_conn,output_list,messenger):
    input_recvd = messenger["input"]
    output_list.append(input_recvd)
    print(input_recvd)
    child_conn.close()

def parent_process(number_of_process=2):
    workers_inputs = [{"input":"hello"}, {"input":"world"}]
    with Manager() as manager:
        processes        = []
        output_list      = manager.list()  # <-- can be shared between processes.
        parent_conn, child_conn = Pipe()
        for single_id_dict in workers_inputs:
            pro_obj             = Process(target=child_process, args=(child_conn,output_list,single_id_dict))  # Passing the list
            pro_obj.start()
            processes.append(pro_obj)
        for p in processes:
            p.join()
        output_list = [single_feature for single_feature in output_list]
    return output_list

parent_process()

OUTPUT:

hello

world

['hello', 'world']

ManagerList is useful to get the output from various parallel process it's like an inbuilt Queue Mechanism with easy to use and safe from deadlocks.