python进程池Pool

需求：

  多进程，加快程序运行。

#示例1
from multiprocessing import Pool, TimeoutError
from multiprocessing import cpu_count
import time
import os

def f(x):
    time.sleep(1)
    return [x*x, x*x], x,'success'

if __name__ == '__main__':
    # start 4 worker processes
    start = time.time()
    pool = Pool(processes=8)
    processTask = []
    aaa = [10,20,30,40]
    for i in range(4):
        processTask.append(pool.apply_async(f, (aaa[i],)))      
    pool.close()
    pool.join()
    print(processTask[0].get()[0])
    print(processTask[1].get()[1])
    print(processTask[2].get()[2])
    print(processTask[3].get())
    print('end1:',time.time()-start)
    
    start = time.time()
    aaa = [10,20,30,40]
    for i in range(4):
        print(i, f(aaa[i]))
    print('end2:',time.time()-start)
    print("CPU的核数为：{}".format(cpu_count()))

结果输出：

[100, 100]
20
success
([1600, 1600], 40, 'success')
end1: 1.7120087146759033
0 ([100, 100], 10, 'success')
1 ([400, 400], 20, 'success')
2 ([900, 900], 30, 'success')
3 ([1600, 1600], 40, 'success')
end2: 4.003762722015381
CPU的核数为：8

———————————————————————————————————————

#示例2
if __name__ == '__main__':
    start = time.time()
    
    aaa = [10,20,30,40]
    processTask = []
    #for k in range(10):
    pool = Pool(processes=4)
    for i in range(4):
        processTask.append(pool.apply_async(f, (aaa[i],)) )
        print(processTask[-1].get())
    pool.close()
    pool.join()
    
    print('end1:',time.time()-start)    
    
    
    start = time.time()
    aaa = [10,20,30,40]
    for i in range(4):
        f(aaa[i])
        #print(i, f(aaa[i]))
    print('end2:',time.time()-start)
    print("CPU的核数为：{}".format(cpu_count()))  

结果输出：

([100, 100], 10, 'success')
([400, 400], 20, 'success')
([900, 900], 30, 'success')
([1600, 1600], 40, 'success')
end1: 4.513966083526611
end2: 4.002805233001709
CPU的核数为：8

综合示例1和2，发现在进程池内部添加返回结果操作会导致程序变慢，每个子进程运行完才执行下一步。实际使用建议使用示例1代码。