how to add specific number of additional workers to an exisiting multiprocessing pool?

问题

In below situation I've created a default pool with two workers and perform tasks. During task processing the task_queue is checked regularly so it doesn't exceeds a certain length limit and prevents up/down stream clutter. How to add dynamically more workers to reduce the task queue length?

import multiprocessing as mp

... code snippet...

def main(poolsize, start_process):

    pool = mp.Pool(processes=poolsize, initializer=start_process)
    done = False

    task_queue = []

    while True:

        ... snippet code : do something ...

        if len(task_queue) >= 10:

            ... code to expand pool goes here...

        if done == True:
            break

    .. do final something ...

if __name__ == '__main__':

#    freeze_support()

    poolsize = 2

    main(poolsize)

回答1:

To add more workers during a running pool processing job you can add below function within the while-loop:

def repopulate(pool, add_workers):

    current_pool_size = len(pool._pool)         # _.pool gets the current pool size.

    new_pool_size = current_pool_size + add_workers

    pool._processes = new_pool_size

    pool._repopulate_pool()

    return pool

Within the while-loop from main():

if len(task_queue) >= 10:

    new_workers = 2

    repopulate(poolname, new_workers)

来源：https://stackoverflow.com/questions/60149376/how-to-add-specific-number-of-additional-workers-to-an-exisiting-multiprocessing