需求：寫一個腳本，判斷192.168.11.0/24網絡里，當前在線ip有哪些？

知識點:

1 使用subprocess模塊，來調用系統命令，執行ping 192.168.11.xxx 命令

2 調用系統命令執行ping命令的時候，會有返回值(ping的結果)，需要用到stdout=fnull, stderr=fnull方法，屏蔽系統執行命令的返回值

常規版本(代碼)

import os
import time
import subprocess
def ping_call():start_time = time.time()fnull = open(os.devnull, 'w')for i in range(1, 256):ipaddr = 'ping 192.168.11.' + str(i)result = subprocess.call(ipaddr + ' -n 2', shell=True, stdout=fnull, stderr=fnull)current_time = time.strftime('%Y%m%d-%H:%M:%S', time.localtime())if result:print('時間:{} ip地址:{} ping fall'.format(current_time, ipaddr))else:print('時間:{} ip地址:{} ping ok'.format(current_time, ipaddr))print('程序耗時{:.2f}'.format(time.time() - start_time))fnull.close()
ping_call()

執行效果：

上面的執行速度非常慢，怎么能讓程序執行速度快起來？

python提供了進程，線程，協程。分別用這三個對上面代碼改進，提高執行效率，測試一波效率

進程池異步執行 – 開啟20個進程

import os
import time
import subprocess
from multiprocessing import Pool
def ping_call(num):fnull = open(os.devnull, 'w')ipaddr = 'ping 192.168.11.' + str(num)result = subprocess.call(ipaddr + ' -n 2', shell=True, stdout=fnull, stderr=fnull)current_time = time.strftime('%Y%m%d-%H:%M:%S', time.localtime())if result:print('時間:{} ip地址:{} ping fall'.format(current_time, ipaddr))else:print('時間:{} ip地址:{} ping ok'.format(current_time, ipaddr))fnull.close()if __name__ == '__main__':start_time = time.time()p = Pool(20)res_l = []for i in range(1, 256):res = p.apply_async(ping_call, args=(i,))res_l.append(res)for res in res_l:res.get()print('程序耗時{:.2f}'.format(time.time() - start_time))

執行結果：

線程池異步執行 – 開啟20個線程

import os
import time
import subprocess
from concurrent.futures import ThreadPoolExecutor
def ping_call(num):fnull = open(os.devnull, 'w')ipaddr = 'ping 192.168.11.' + str(num)result = subprocess.call(ipaddr + ' -n 2', shell=True, stdout=fnull, stderr=fnull)current_time = time.strftime('%Y%m%d-%H:%M:%S', time.localtime())if result:print('時間:{} ip地址:{} ping fall'.format(current_time, ipaddr))else:print('時間:{} ip地址:{} ping ok'.format(current_time, ipaddr))fnull.close()if __name__ == '__main__':start_time = time.time()thread_pool = ThreadPoolExecutor(20)ret_lst = []for i in range(1, 256):ret = thread_pool.submit(ping_call, i)ret_lst.append(ret)thread_pool.shutdown()for ret in ret_lst:ret.result()print('線程池(20)異步-->耗時{:.2f}'.format(time.time() - start_time))

執行結果：

協程執行—(執行多個任務，遇到I/O操作就切換)

使用gevent前，需要pip install gevent

from gevent import monkey;monkey.patch_all()
import gevent
import os
import time
import subprocessdef ping_call(num):fnull = open(os.devnull, 'w')ipaddr = 'ping 192.168.11.' + str(num)result = subprocess.call(ipaddr + ' -n 2', shell=True, stdout=fnull, stderr=fnull)current_time = time.strftime('%Y%m%d-%H:%M:%S', time.localtime())if result:print('時間:{} ip地址:{} ping fall'.format(current_time, ipaddr))else:print('時間:{} ip地址:{} ping ok'.format(current_time, ipaddr))fnull.close()def asynchronous(): # 異步g_l = [gevent.spawn(ping_call, i) for i in range(1, 256)]gevent.joinall(g_l)if __name__ == '__main__':start_time = time.time()asynchronous()print('協程執行-->耗時{:.2f}'.format(time.time() - start_time))

執行結果：

遇到I/O操作，協程的效率比進程，線程高很多!

總結：python中，涉及到I/O阻塞的程序中，使用協程的效率最高

最后附帶協程池代碼

gevent.pool

from gevent import monkey;monkey.patch_all()
import gevent
import os
import time
import subprocess
import gevent.pooldef ping_call(num):fnull = open(os.devnull, 'w')ipaddr = 'ping 192.168.11.' + str(num)result = subprocess.call(ipaddr + ' -n 2', shell=True, stdout=fnull, stderr=fnull)current_time = time.strftime('%Y%m%d-%H:%M:%S', time.localtime())if result:print('時間:{} ip地址:{} ping fall'.format(current_time, ipaddr))else:print('時間:{} ip地址:{} ping ok'.format(current_time, ipaddr))fnull.close()if __name__ == '__main__':start_time = time.time()res_l = []p = gevent.pool.Pool(100)for i in range(1, 256):res_l.append(p.spawn(ping_call, i))gevent.joinall(res_l)print('協程池執行-->耗時{:.2f}'.format(time.time() - start_time))

執行結果: