ApacheBench is a command line computer program for measuring the performance of HTTP web servers, in particular the Apache HTTP Server. It was designed to give an idea of the performance that a given Apache installation can provide. In particular, it shows how many requests per second the server is capable of serving.
You can provide it with cookie/value pairs of a logged in user, and point it to the search URL, with the ability to tell it to run concurrent connections to simulate concurrent users.
Apache JMeter would definitely support your scenario and supports Distributed Testing (pdf) which is a mode where you run several agents on different machines and can thus scale to simulate high numbers of concurrent users.
You could try Web Performance Load Tester - it is a mid-priced load testing product. It is much less expensive than SilkPerformer or LoadRunner and much more productive. It is limited to only web protocols, but that doesn't sound like it will be a problem for your application. There are videos on the site that will give you a pretty good feel for the tool: http://www.webperformanceinc.com/load_testing/demo/
(disclaimer: I work for Web Performance and use the product nearly every day)
I recommend you to try the Microsoft Web Capacity Analysis Tool (WCAT). It is able to run on multiple machines simultaneously, and each of the clients can have multiple "virtual clients", thus they are able to simulate multiple connections to a web server. It can connect to any web server (used by some Apache fanatics), but of course it integrates best with IIS.
It was developed by the IIS performance team and is used by Microsoft to stress-test their IIS server. It has both x86 and x64 versions, and it is easy to create your own extensions to it.
It works the following way: you create a scenario file with which you can describe any number of HTTP requests. (It has options for GET, POST, and also has HTTPS support.) You can then group these requests to transactions which represent a series of requests to simulate user interaction. Each virtual client runs a single transaction at a time, and it waits until the current request completed before starting the next one. It is easy to write a transaction that contains a series of requests which can simulate users "clicking" on things. (It is also possible to "sleep" for a certain amount of time between requests, thus simulating a user reading the page.)
Here is the official introduction:
Web Capacity Analysis Tool (WCAT) is a lightweight HTTP load generation tool primarily designed to measure the performance of a web server within a controlled environment. WCAT can simulate thousands of concurrent users making requests to a single web site or multiple web sites. The WCAT engine uses a simple script to define the set of HTTP requests to be played back to the web server. Extensibility is provided through plug-in DLLs and a standard, simple API.
Perhaps its only downside is that it is a bit compicated to set up at first. It is a command line tool so it is also a bit too easy to screw up with the parameters.
What is the financial risk if the application does not work/scale?
What is the financial value of the output of the performance tests in deciding to go live or tune|modify the existing application?
How does the financial impact and the value of the output of the tests match to your cost of tool issue?
Generally speaking if the risk is high enough to merit a performance test then the value of the output of a solid performance test will dwarf the cost of any tool deployed to conduct the test. The caveat here is that you need to be willing to bring on board a solid performance engineer to design and run the tests, for even the least expensive test tools can run a substantial negative ROI with an insufficiently skilled person at the healm ( and it happens all the time ).
发布评论
评论（8）
使用 apache 的基准测试工具
ab。来自 维基百科：ApacheBench：
您可以向它提供登录用户的 cookie/值对，并将其指向搜索 URL，并能够告诉它运行并发连接来模拟并发用户。
Use apache's benchmark tool
ab.From Wikipedia: ApacheBench:
You can provide it with cookie/value pairs of a logged in user, and point it to the search URL, with the ability to tell it to run concurrent connections to simulate concurrent users.
Apache JMeter 肯定会支持您的场景并支持 分布式测试 (pdf)，这是一种在不同计算机上运行多个代理的模式，因此可以扩展以模拟大量并发用户。
Apache JMeter would definitely support your scenario and supports Distributed Testing (pdf) which is a mode where you run several agents on different machines and can thus scale to simulate high numbers of concurrent users.
您可以尝试 Web Performance Load Tester - 它是一款中等价格的负载测试产品。它比 SilkPerformer 或 LoadRunner 便宜得多，而且生产力更高。它仅限于网络协议，但这听起来不会对您的应用程序造成问题。网站上有一些视频可以让您很好地体验该工具：http://www .webperformanceinc.com/load_testing/demo/
（免责声明：我为 Web Performance 工作，几乎每天都使用该产品）
You could try Web Performance Load Tester - it is a mid-priced load testing product. It is much less expensive than SilkPerformer or LoadRunner and much more productive. It is limited to only web protocols, but that doesn't sound like it will be a problem for your application. There are videos on the site that will give you a pretty good feel for the tool: http://www.webperformanceinc.com/load_testing/demo/
(disclaimer: I work for Web Performance and use the product nearly every day)
还有 httperf：
http://agiletesting.blogspot。 com/2005/04/http-performance-testing-with-httperf.html
There's also httperf:
http://agiletesting.blogspot.com/2005/04/http-performance-testing-with-httperf.html
我建议您尝试 Microsoft Web 容量分析工具（WCAT）。它能够同时在多台机器上运行，并且每个客户端都可以有多个“虚拟客户端”，因此它们能够模拟与Web服务器的多个连接。它可以连接到任何 Web 服务器（由一些 Apache 狂热者使用），但当然它与 IIS 集成得最好。
它由 IIS 性能团队开发，并被 Microsoft 用于对其 IIS 服务器进行压力测试。它有 x86 和 x64 版本，并且可以轻松创建您自己的扩展。
它的工作方式如下：您创建一个场景文件，用它可以描述任意数量的 HTTP 请求。 （它具有 GET、POST 选项，并且还支持 HTTPS。）然后，您可以将这些请求分组到事务中，这些事务代表一系列模拟用户交互的请求。每个虚拟客户端一次运行一个事务，并等到当前请求完成后才开始下一个请求。
编写包含一系列请求的事务很容易，这些请求可以模拟用户“单击”事物。 （也可以在请求之间“休眠”一定的时间，从而模拟用户阅读页面。）
下面是官方的介绍：
也许它唯一的缺点是一开始设置有点复杂。它是一个命令行工具，因此也很容易弄乱参数。
I recommend you to try the Microsoft Web Capacity Analysis Tool (WCAT). It is able to run on multiple machines simultaneously, and each of the clients can have multiple "virtual clients", thus they are able to simulate multiple connections to a web server. It can connect to any web server (used by some Apache fanatics), but of course it integrates best with IIS.
It was developed by the IIS performance team and is used by Microsoft to stress-test their IIS server. It has both x86 and x64 versions, and it is easy to create your own extensions to it.
It works the following way: you create a scenario file with which you can describe any number of HTTP requests. (It has options for GET, POST, and also has HTTPS support.) You can then group these requests to transactions which represent a series of requests to simulate user interaction. Each virtual client runs a single transaction at a time, and it waits until the current request completed before starting the next one.
It is easy to write a transaction that contains a series of requests which can simulate users "clicking" on things. (It is also possible to "sleep" for a certain amount of time between requests, thus simulating a user reading the page.)
Here is the official introduction:
Perhaps its only downside is that it is a bit compicated to set up at first. It is a command line tool so it is also a bit too easy to screw up with the parameters.
对于复杂的网站，您可以使用 WebLOAD。它不是免费的，但也不是太贵。
For complex sites you can use WebLOAD. It's not free, but not too expensive either.
如果应用程序无法运行/无法扩展，财务风险是什么？
在决定上线或调整|修改现有应用程序时，性能测试输出的财务价值是什么？< /strong>
测试输出的财务影响和价值与您的工具问题成本如何匹配？
一般来说，如果风险足够高，值得进行性能测试，那么测试结果的价值如何？可靠的性能测试的输出将使部署用于执行测试的任何工具的成本相形见绌。这里需要注意的是，您需要愿意聘请一位可靠的性能工程师来设计和运行测试，因为即使是最便宜的测试工具也可能会在技术人员不够熟练的情况下产生显着的负投资回报率（并且这种情况发生了）一直）。
What is the financial risk if the application does not work/scale?
What is the financial value of the output of the performance tests in deciding to go live or tune|modify the existing application?
How does the financial impact and the value of the output of the tests match to your cost of tool issue?
Generally speaking if the risk is high enough to merit a performance test then the value of the output of a solid performance test will dwarf the cost of any tool deployed to conduct the test. The caveat here is that you need to be willing to bring on board a solid performance engineer to design and run the tests, for even the least expensive test tools can run a substantial negative ROI with an insufficiently skilled person at the healm ( and it happens all the time ).
Java 应用程序的磨床：Grinder
the grinder for java apps : Grinder