实例讲解Python中的多线程、多进程如何应对IO密集型任务、计算密集型任务

这里通过一个实例，说明多线程适合IO密集型任务，多进程适合计算密集型任务。
一、IO密集型任务：

import multiprocessing
import time
import threading 
# 定义全局变量Queue
g_queue = multiprocessing.Queue()
 
# 首先定义一个队列，并定义初始化队列的函数：
def init_queue():
    print("init g_queue start")
    while not g_queue.empty():
        g_queue.get()
    for _index in range(10):
        g_queue.put(_index)
    print("init g_queue end")
    return

# 定义一个IO密集型任务：利用time.sleep()，分别从队列中获取任务数据
def task_io(task_id):
    print("IOTask[%s] start" % task_id)
    while not g_queue.empty():
        time.sleep(1)
        try:
            data = g_queue.get(block=True, timeout=1)
            print("IOTask[%s] get data: %s" % (task_id, data))
        except Exception as excep:
            print("IOTask[%s] error: %s" % (task_id, str(excep)))
    print("IOTask[%s] end" % task_id)
    return
 
if __name__ == '__main__':
    print("cpu count:", multiprocessing.cpu_count(), "\n")
 
    print("========== 直接执行IO密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    task_io(0)
    print("结束：", time.time() - time_0, "\n")
 
    print("========== 多线程执行IO密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    thread_list = [threading.Thread(target=task_io, args=(i,)) for i in range(5)]
    for t in thread_list:
        t.start()
    for t in thread_list:
        if t.is_alive():
            t.join()
    print("结束：", time.time() - time_0, "\n")
 
    print("========== 多进程执行IO密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    process_list = [multiprocessing.Process(target=task_io, args=(i,)) for i in range(multiprocessing.cpu_count())]
    for p in process_list:
        p.start()
    for p in process_list:
        if p.is_alive():
            p.join()
    print("结束：", time.time() - time_0, "\n")

打印：

[root@localhost demo01]# python3 text1109.py 
cpu count: 4 

========== 直接执行IO密集型任务 ==========
init g_queue start
init g_queue end
IOTask[0] start
IOTask[0] get data: 0
IOTask[0] get data: 1
IOTask[0] get data: 2
IOTask[0] get data: 3
IOTask[0] get data: 4
IOTask[0] get data: 5
IOTask[0] get data: 6
IOTask[0] get data: 7
IOTask[0] get data: 8
IOTask[0] get data: 9
IOTask[0] end
结束： 10.016204833984375 

========== 多线程执行IO密集型任务 ==========
init g_queue start
init g_queue end
IOTask[0] start
IOTask[0] end
IOTask[1] start
IOTask[3] start
IOTask[2] start
IOTask[4] start
IOTask[1] get data: 0
IOTask[2] get data: 1
IOTask[3] get data: 2
IOTask[4] get data: 3
IOTask[1] get data: 4
IOTask[2] get data: 5
IOTask[4] get data: 6
IOTask[3] get data: 7
IOTask[1] get data: 8
IOTask[2] get data: 9
IOTask[2] end
IOTask[3] error: 
IOTask[3] end
IOTask[4] error: 
IOTask[4] end
IOTask[1] error: 
IOTask[1] end
结束： 5.012259006500244 

========== 多进程执行IO密集型任务 ==========
init g_queue start
init g_queue end
IOTask[0] start
IOTask[1] start
IOTask[2] start
IOTask[3] start
IOTask[0] get data: 0
IOTask[1] get data: 1
IOTask[2] get data: 2
IOTask[3] get data: 3
IOTask[0] get data: 4
IOTask[1] get data: 5
IOTask[2] get data: 6
IOTask[3] get data: 7
IOTask[0] get data: 8
IOTask[1] get data: 9
IOTask[1] end
IOTask[3] error: 
IOTask[3] end
IOTask[2] error: 
IOTask[2] end
IOTask[0] error: 
IOTask[0] end
结束： 5.0164711475372314 

结果说明：

对于IO密集型任务：

• 直接执行用时：10.016204833984375秒
• 多线程执行用时：5.012259006500244秒
• 多进程执行用时：5.0164711475372314秒

说明多线程适合IO密集型任务。

二、计算密集型任务：

import multiprocessing
import time
import threading 
# 定义全局变量Queue
g_queue = multiprocessing.Queue()
 
def init_queue():
    print("init g_queue start")
    while not g_queue.empty():
        g_queue.get()
    for _index in range(10):
        g_queue.put(_index)
    print("init g_queue end")
    return

g_search_list = list(range(10000))
# 定义一个计算密集型任务：利用一些复杂加减乘除、列表查找等
def task_cpu(task_id):
    print("CPUTask[%s] start" % task_id)
    while not g_queue.empty():
        count = 0
        for i in range(10000):
            count += pow(3*2, 3*2) if i in g_search_list else 0
        try:
            data = g_queue.get(block=True, timeout=1)
            print("CPUTask[%s] get data: %s" % (task_id, data))
        except Exception as excep:
            print("CPUTask[%s] error: %s" % (task_id, str(excep)))
    print("CPUTask[%s] end" % task_id)
    return task_id

if __name__ == '__main__':
    print("cpu count:", multiprocessing.cpu_count(), "\n")
 
    print("========== 直接执行CPU密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    task_cpu(0)
    print("结束：", time.time() - time_0, "\n")
 
    print("========== 多线程执行CPU密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    thread_list = [threading.Thread(target=task_cpu, args=(i,)) for i in range(5)]
    for t in thread_list:
        t.start()
    for t in thread_list:
        if t.is_alive():
            t.join()
    print("结束：", time.time() - time_0, "\n")
 
    print("========== 多进程执行cpu密集型任务 ==========")
    init_queue()
    time_0 = time.time()
    process_list = [multiprocessing.Process(target=task_cpu, args=(i,)) for i in range(multiprocessing.cpu_count())]
    for p in process_list:
        p.start()
    for p in process_list:
        if p.is_alive():
            p.join()
    print("结束：", time.time() - time_0, "\n")

打印：

[root@localhost demo01]# python3 text1109.py 
cpu count: 4 

========== 直接执行CPU密集型任务 ==========
init g_queue start
init g_queue end
CPUTask[0] start
CPUTask[0] get data: 0
CPUTask[0] get data: 1
CPUTask[0] get data: 2
CPUTask[0] get data: 3
CPUTask[0] get data: 4
CPUTask[0] get data: 5
CPUTask[0] get data: 6
CPUTask[0] get data: 7
CPUTask[0] get data: 8
CPUTask[0] get data: 9
CPUTask[0] end
结束： 5.545855522155762 

========== 多线程执行CPU密集型任务 ==========
init g_queue start
init g_queue end
CPUTask[0] start
CPUTask[1] start
CPUTask[2] start
CPUTask[4] start
CPUTask[3] start
CPUTask[0] get data: 0
CPUTask[2] get data: 1
CPUTask[1] get data: 2
CPUTask[4] get data: 3
CPUTask[3] get data: 4
CPUTask[0] get data: 5
CPUTask[2] get data: 6
CPUTask[1] get data: 7
CPUTask[3] get data: 8
CPUTask[4] get data: 9
CPUTask[4] end
CPUTask[2] error: 
CPUTask[2] end
CPUTask[1] error: 
CPUTask[1] end
CPUTask[0] error: 
CPUTask[0] end
CPUTask[3] error: 
CPUTask[3] end
结束： 9.969066858291626 

========== 多进程执行cpu密集型任务 ==========
init g_queue start
init g_queue end
CPUTask[0] start
CPUTask[1] start
CPUTask[3] start
CPUTask[2] start
CPUTask[0] get data: 0
CPUTask[1] get data: 1
CPUTask[3] get data: 2
CPUTask[2] get data: 3
CPUTask[0] get data: 4
CPUTask[1] get data: 5
CPUTask[3] get data: 6
CPUTask[2] get data: 7
CPUTask[0] get data: 8
CPUTask[1] get data: 9
CPUTask[1] end
CPUTask[3] error: 
CPUTask[3] end
CPUTask[2] error: 
CPUTask[2] end
CPUTask[0] error: 
CPUTask[0] end
结束： 3.3789467811584473 

结果说明：

对于计算密集型任务：

• 直接执行用时：5.545855522155762秒
• 多线程执行用时：9.969066858291626秒
• 多进程执行用时：3.3789467811584473秒

说明多进程适合计算密集型任务。

参考链接：
Python进阶：聊聊IO密集型任务、计算密集型任务，以及多线程、多进程
进程通信(multiprocessing.Queue)