Question

使用Python 3的多处理Pool，如何获得工作进程当前执行的作业。在下面的简化示例中，我还将变量x定义为特定进程的ID，这正是我要寻找的。

from multiprocessing import Pool

done_tasks = []

def f(x):
    return x*x

def on_task_done(x):
    done_tasks.append(x)

if __name__ == '__main__':
    pool = multiprocessing.Pool(processes=3)
    ids = [0, 1, 2, 3, 4, ...]
    for id in ids:
        pool.apply_async(start_worker, id, callback=on_task_done)

    # wait and print currently running processes
    while True:
        # TODO print currently running processes

        if len(done_tasks) >= len(ids):
            break

具体来说，如何找到工作进程在任何给定时间点当前正在执行的进程的ID（例如，在循环等待直到提交给池的所有作业完成时）？

Answer 1

您可以使用Manager实例创建一个dict，该实例可用于存储当前正在运行的作业的ID。这是您可以执行的操作：

import uuid
import time
import random
from multiprocessing import Pool, Manager


done_tasks = []


def square(x, procs):
    pid = uuid.uuid4().hex
    procs[pid] = True

    # Simulate a long-running task:
    time.sleep(random.randint(1, 3))

    procs[pid] = False

    return x * x


def on_task_done(results):
    done_tasks.append(results)


def main():
    m = Manager()
    procs = m.dict()
    pool = Pool(processes=2)

    for x in range(10):
        pool.apply_async(square, args=(x, procs), callback=on_task_done)

    while len(done_tasks) < 10:
        pids = [pid for pid, running in procs.items() if running]
        print('running jobs:', pids)
        time.sleep(1)

    print('results:', done_tasks)

    pool.close()
    pool.join()


if __name__ == '__main__':
    main()

输出：

running jobs: []
running jobs: ['949784a09e1d4f09b647cae584e86277', 'de3c23dd3c174e9a8d7f5c2a7d89a183']
running jobs: ['de3c23dd3c174e9a8d7f5c2a7d89a183']
running jobs: ['47f536e9e3584e808bd7475ad43ee49a']
running jobs: ['e06d73341d774f86a8a00d0bb2914642', '874b194b548e4f47b02859b4f704a4f2']
running jobs: ['874b194b548e4f47b02859b4f704a4f2', 'cb95a6bcc8ff47bba271cb845a919245']
running jobs: ['3c9fbf7e4c604dedae033b7ef93c3523', '4863c18062504a8190b415bc42874bb7']
running jobs: ['3c9fbf7e4c604dedae033b7ef93c3523', '4863c18062504a8190b415bc42874bb7']
running jobs: ['3c9fbf7e4c604dedae033b7ef93c3523', '4863c18062504a8190b415bc42874bb7']
running jobs: ['2f896b86aa1647fea63c522fb2c67684', 'aba21db1b1af4fd8896a6b59534a254f']
running jobs: ['2f896b86aa1647fea63c522fb2c67684']
running jobs: ['2f896b86aa1647fea63c522fb2c67684']
results: [0, 1, 4, 9, 16, 25, 36, 49, 81, 64]

Python多处理池：如何使当前所有进程运行

1 个答案: