Do you have access to Python running on Python? Other processes query processes that take a long time to initialize and receive results.

Yes, it's possible.

I had a similar experience.

In my case, when I received a user's request from the WebServer, I executed the corresponding Python code and sent a response. However, it took a while for the Python code to go up on memory.

It happened because the code was executed only when there was a request, and all of the memory was deleted after writing Response.

The IPC concept was needed to solve this problem. InterProcess Communication can also share memory, but it is a technique used to maximize performance in places like C and Cpp. It can be solved like that, but the general solution is to use Socket a lot for IPC. I mean, the Server-Client model.

The problematic part should be returned to the server application at all times to the Damon state should be maintained.

It's the code I used before. It's an old code, so it might not fit right now.

Please keep that in mind ^.

from multiprocessing.connection import Listener
from konlpy.tag import Kkma

from array import array

address = ('localhost', 65123)
password = b'1234'
listener = Listener(address, authkey=password)

#initialize
if __name__ == '__main__':
    kkma = Kkma()
    kkma.nouns('initializing')
    print('main!')

while True:
    conn = listener.accept()
    print('connection accepted from', listener.last_accepted)
    msg = conn.recv()
    sendMessage = kkma.nouns(msg)
    print(sendMessage)
    conn.send(sendMessage)
    conn.close()