如何在Python中调用不同的URL进行负载测试？

我有一个针对任何特定url执行负载测试的代码。但我必须对具有不同URL的web服务进行负载测试。为此，我需要创建一个URL数组，每个线程都应该命中数组中给定的所有URL。我该怎么做？这是我的代码：

import httplib2

import socket

import time

from threading import Event
from threading import Thread
from threading import current_thread
from urllib import urlencode

# Modify these values to control how the testing is done

# How many threads should be running at peak load.
NUM_THREADS = 50

# How many minutes the test should run with all threads active.
TIME_AT_PEAK_QPS = 20 # minutes

# How many seconds to wait between starting threads.
# Shouldn't be set below 30 seconds.
DELAY_BETWEEN_THREAD_START = 30 # seconds

quitevent = Event()

def threadproc():
    """This function is executed by each thread."""
    print "Thread started: %s" % current_thread().getName()
    h = httplib2.Http(timeout=30)
    while not quitevent.is_set():
        try:
            # HTTP requests to exercise the server go here
            # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
            resp, content = h.request(
                "http://www.google.com")
            if resp.status != 200:
             print "Response not OK"
            # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
        except socket.timeout:
            pass

print "Thread finished: %s" % current_thread().getName()


if __name__ == "__main__":
    runtime = (TIME_AT_PEAK_QPS * 60 + DELAY_BETWEEN_THREAD_START * NUM_THREADS)
    print "Total runtime will be: %d seconds" % runtime
    threads = []
    try:
        for i in range(NUM_THREADS):
            t = Thread(target=threadproc)
            t.start()
            threads.append(t)
            time.sleep(DELAY_BETWEEN_THREAD_START)
        print "All threads running"
        time.sleep(TIME_AT_PEAK_QPS*60)
        print "Completed full time at peak qps, shutting down threads"
    except:
        print "Exception raised, shutting down threads"
quitevent.set()
time.sleep(3)
for t in threads:
    t.join(1.0)
print "Finished"

---

扩展类，而不是将

threadproc

传递给

Thread

：

class Worker(Thread):
    def __init__(self, urls):
        super(Worker, self).__init__()
        self.urls = urls

    def run(self):
        for url in self.urls:
            self.fetch(url)

---

也就是说，除非您这样做是为了更好地理解线程和负载测试如何在内部工作，否则我建议您使用一个成熟的测试框架，比如。多年的经验积累了下来，你必须先积累起来。

你还没有解释你所拥有的东西有什么问题。你有什么错误吗？结果是否与预期不符？如果您真的要对服务器进行负载测试，我建议您使用专门为此目的设计的工具。我成功使用的一个是开源ApacheJMeter（）。我还建议使用tsung进行负载测试（）。事实上，我是python的初学者，您解释的类很容易理解，但我不知道在代码中我必须提到针对该负载测试执行的所有URL。您只能看到code resp中提到的一个url，content=h.request（“），但我如何将其分配给多个url？将

threadproc

移动到新类

Worker

，为其分配参数

self，url

，然后将

h.request（“google.com”）

替换为

h.request（url）

Thanx，但我必须在其中定义URL数组。例如example.com、example.com/1、example.com/2。。。。