Reason
In many cases, we write the process only to the fixed module to take the concurrency limit, and the ASP also has his own processing mechanism, especially for the focus on business, results-oriented team may pay more attention to a certain time node to give a satisfactory product, but your module in the end is robust, you have a mental spectrum? Tell me aloud is you sure? Let's look through a tool. He is jmeter!
(The final test found that the blog park visit some problems, the specific hope that the majority of the park friends to give guidance)
Basic IntroductionApache JMeter is a Java-based stress testing tool developed by the Apache organization. For stress testing of software, it was originally designed for Web application testing but later extended to other test areas. It can be used to test static and dynamic resources such as static files, Java applets, CGI scripts, Java objects, databases, FTP servers, and so on. JMeter can be used to simulate huge loads on servers, networks, or objects, testing their strength and analyzing overall performance from different pressure classes. In addition, JMeter is able to perform functional/regression testing of the application by creating a script with assertions to verify that your program returns the results you expect. For maximum flexibility, JMeter allows you to create assertions using regular expressions.
Environment preparation
(1) JMeter is built in the Java environment, so your machine needs to install the Java environment, I like the Java Group of colleagues to a brief installation package, after installation needs to configure the environment variables.
(2) JMeter installation Package http://jmeter.apache.org/
Because the blog park on the upload file has a size limit, so ...
Java environment:
Jmeter:
Basic Concept Understanding
Test Plan
Thread Group
Each task in the test has to be thread-processed, and all our subsequent tasks must be created under the Threads group.
Logic Control Unit
Many different implementations can determine the order in which each sample is executed.
Sampling device
It can be assumed that all test tasks are taken by the sampler, and there are some, such as: HTTP requests.
Timer
The timer is used to set the wait time between operations, and the wait time is the hand side of the performance test commonly used to control the client QPS. Similar to the "Think Time" inside the LoadRunner. JMeter defines different types of timers, such as bean Shell Timer, Constant throughput timer, fixed timer, and so on.
Assertion
The result of the request returned by the sampler is judged correctly.
Configuring components
The configuration element (config Element) is used to provide support for static data configuration. The CSV data Set config can form a data pool for the local data file, while the configuration for the types of HTTP request sampler and TCP Request Sampler can modify the default data for sampler. (for example, HTTP cookie manager can be used to manage HTTP Request Sampler cookies)
Listener
Its function is to display the request result of sampler, statistic some data (throughput, kb/s ...) such as
Getting Started
Let's take a look at some examples of jmeter
Test Address: http://www.cnblogs.com/mongo/p/4623560.html
Scenario: Simulate user concurrent requests to view Web site load Parameters
(1) Run JMeter, install the Java environment, open the JMeter installation directory, run \apache-jmeter-2.13\bin\jmeter.bat
(2) Add thread Group
Description
Number of threads: We've configured 500 threads open
Ramp-up Period: Set to 0 indicates that 500 threads are started at the same time within 0 seconds
Cycle times: Overall execution once
(3) Setting the controller
Here we set the loop controller, each thread loops 5 times Sampler Operation
(4) Set Sampler Sampler
Here we use a blog park article address to do test, simple request page
(5) Add monitoring listener
In order to better see our test results, we need the appropriate graphical display, where we add graphical results and observation results tree
(6) Operating Results (6.1) Summary Report
Number of requests 2500
Average response Time 1185
Throughput: 175.8/s
(6.2) Graphical results
(6.3) Observation result tree
There seems to be a problem here, I visit the blog Park designated pages, the previous requests can be, and then the request has been time out state, over a period request again, there are currently 2 conjecture
(1) The blog Park has made restrictions, a certain address frequently send requests will be limited to a period of time?
(2) But obviously the number of successful start requests and the number of requests after a period of time is obviously different, is there really a problem with the program?
This is just speculation, specific people can also consider the exchange of guidance, this article is mainly to allow you to test a certain part of the load, introduced the basic use of jmeter, so that we have a quantitative understanding of their own modules, in the writing process will be more thinking.
Performance Test-jmeter