Python中单线程、多线程和多进程的效率对比实验
HenryAlarco
7年前
<p>Python是运行在解释器中的语言，查找资料知道，python中有一个全局锁（GIL），在使用多进程(Thread)的情况下，不能发挥多核的优势。而使用多进程(Multiprocess)，则可以发挥多核的优势真正地提高效率。</p> <h2><strong>对比实验</strong></h2> <p>资料显示，如果多线程的进程是 <strong>CPU密集型</strong> 的，那多线程并不能有多少效率上的提升，相反还可能会因为线程的频繁切换，导致效率下降，推荐使用多进程；如果是 <strong>IO密集型</strong> ，多线程进程可以利用IO阻塞等待时的空闲时间执行其他线程，提升效率。所以我们根据实验对比不同场景的效率</p> <table> <thead> <tr> <th>操作系统</th> <th>CPU</th> <th>内存</th> <th>硬盘</th> </tr> </thead> <tbody> <tr> <td>Windows 10</td> <td>双核</td> <td>8GB</td> <td>机械硬盘</td> </tr> </tbody> </table> <p>(1)引入所需要的模块</p> <pre> <code class="language-python">importrequests importtime fromthreadingimportThread frommultiprocessingimportProcess </code></pre> <p>(2)定义CPU密集的计算函数</p> <pre> <code class="language-python">def count(x, y): # 使程序完成150万计算 c = 0 while c < 500000: c += 1 x += x y += y </code></pre> <p>（3）定义IO密集的文件读写函数</p> <pre> <code class="language-python">defwrite(): f = open("test.txt", "w") for x in range(5000000): f.write("testwrite\n") f.close() defread(): f = open("test.txt", "r") lines = f.readlines() f.close() </code></pre> <p>(4) 定义网络请求函数</p> <pre> <code class="language-python">_head = { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.116 Safari/537.36'} url = "http://www.tieba.com" defhttp_request(): try: webPage = requests.get(url, headers=_head) html = webPage.text return {"context": html} exceptExceptionas e: return {"error": e} </code></pre> <p>(5)测试线性执行IO密集操作、CPU密集操作所需时间、网络请求密集型操作所需时间</p> <pre> <code class="language-python"># CPU密集操作 t = time.time() for x in range(10): count(1, 1) print("Line cpu", time.time() - t) # IO密集操作 t = time.time() for x in range(10): write() read() print("Line IO", time.time() - t) # 网络请求密集型操作 t = time.time() for x in range(10): http_request() print("Line Http Request", time.time() - t) </code></pre> <p>输出</p> <ul> <li>CPU密集：95.6059999466、91.57099986076355 92.52800011634827、 99.96799993515015</li> <li>IO密集：24.25、21.76699995994568、21.769999980926514、22.060999870300293</li> <li>网络请求密集型: 4.519999980926514、8.563999891281128、4.371000051498413、4.522000074386597、14.671000003814697</li> </ul> <p>(6)测试多线程并发执行CPU密集操作所需时间</p> <pre> <code class="language-python">counts = [] t = time.time() for x in range(10): thread = Thread(target=count, args=(1,1)) counts.append(thread) thread.start() e = counts.__len__() while True: for thin counts: if not th.is_alive(): e -= 1 if e <= 0: break print(time.time() - t) </code></pre> <p>Output: 99.9240000248 、101.26400017738342、102.32200002670288</p> <p>(7)测试多线程并发执行IO密集操作所需时间</p> <pre> <code class="language-python">def io(): write() read() t = time.time() ios = [] t = time.time() for x in range(10): thread = Thread(target=count, args=(1,1)) ios.append(thread) thread.start() e = ios.__len__() while True: for thin ios: if not th.is_alive(): e -= 1 if e <= 0: break print(time.time() - t) </code></pre> <p>Output: 25.69700002670288、24.02400016784668</p> <p>(8)测试多线程并发执行网络密集操作所需时间</p> <pre> <code class="language-python">t = time.time() ios = [] t = time.time() for x in range(10): thread = Thread(target=http_request) ios.append(thread) thread.start() e = ios.__len__() while True: for thin ios: if not th.is_alive(): e -= 1 if e <= 0: break print("Thread Http Request", time.time() - t) </code></pre> <p>Output: 0.7419998645782471、0.3839998245239258、0.3900001049041748</p> <p>(9)测试多进程并发执行CPU密集操作所需时间</p> <pre> <code class="language-python">counts = [] t = time.time() for x in range(10): process = Process(target=count, args=(1,1)) counts.append(process) process.start() e = counts.__len__() while True: for thin counts: if not th.is_alive(): e -= 1 if e <= 0: break print("Multiprocess cpu", time.time() - t) </code></pre> <p>Output: 54.342000007629395、53.437999963760376</p> <p>(10)测试多进程并发执行IO密集型操作</p> <pre> <code class="language-python">t = time.time() ios = [] t = time.time() for x in range(10): process = Process(target=io) ios.append(process) process.start() e = ios.__len__() while True: for thin ios: if not th.is_alive(): e -= 1 if e <= 0: break print("Multiprocess IO", time.time() - t) </code></pre> <p>Output: 12.509000062942505、13.059000015258789</p> <p>(11)测试多进程并发执行Http请求密集型操作</p> <pre> <code class="language-python">t = time.time() httprs = [] t = time.time() for x in range(10): process = Process(target=http_request) ios.append(process) process.start() e = httprs.__len__() while True: for thin httprs: if not th.is_alive(): e -= 1 if e <= 0: break print("Multiprocess Http Request", time.time() - t) </code></pre> <p>Output: 0.5329999923706055、0.4760000705718994</p> <h2><strong>实验结果</strong></h2> <table> <thead> <tr> <th> </th> <th>CPU密集型操作</th> <th>IO密集型操作</th> <th>网络请求密集型操作</th> </tr> </thead> <tbody> <tr> <td>线性操作</td> <td>94.91824996469</td> <td>22.46199995279</td> <td>7.3296000004</td> </tr> <tr> <td>多线程操作</td> <td>101.1700000762</td> <td>24.8605000973</td> <td>0.5053332647</td> </tr> <tr> <td>多进程操作</td> <td>53.8899999857</td> <td>12.7840000391</td> <td>0.5045000315</td> </tr> </tbody> </table> <p>通过上面的结果，我们可以看到：</p> <ul> <li>多线程在IO密集型的操作下似乎也没有很大的优势（也许IO操作的任务再繁重一些就能体现出优势），在CPU密集型的操作下明显地比单线程线性执行性能更差，但是对于网络请求这种忙等阻塞线程的操作，多线程的优势便非常显著了</li> <li>多进程无论是在CPU密集型还是IO密集型以及网络请求密集型（经常发生线程阻塞的操作）中，都能体现出性能的优势。不过在类似网络请求密集型的操作上，与多线程相差无几，但却更占用CPU等资源，所以对于这种情况下，我们可以选择多线程来执行</li> </ul> <p style="text-align:center"><br> <img src="https://simg.open-open.com/show/2ee3910aaab0c8c49ffd7bcaf2037493.png"></p> <p> </p> <p>来自：http://python.jobbole.com/86822/</p> <p> </p>