xdocs/usermanual/jmeter_proxy_step_by

<?xml version="1.0"?>  <!DOCTYPE document[ <!ENTITY sect-num '26'> <!ENTITY hellip "…" > ]> <document prev="jmeter_distributed_testing_step_by_step.html" next="junitsampler_tutorial.html" id="$Id$"> <properties> <author email="dev@jmeter.apache.org">JMeter developers</author> <title>Apache JMeter HTTP(S) Test Script Recorder</title> </properties> <body> <section name="&sect-num;. Apache JMeter HTTP(S) Test Script Recorder" anchor="script-recorder"> This tutorial attempts to explain the exact steps for recording HTTP/HTTPS. For those new to JMeter, one easy way to create a test plan is to use the Recorder. <subsection name="&sect-num;.1 JMeter configuration" anchor="configuration"> Since JMeter 2.10, recording has been improved to better handle embedded resources and creation of certificates on the fly. To enable these features, JMeter uses <code>keytool</code> utility (available in JRE/JDK) so you need to ensure your configuration is correct, read <a href="https://cwiki.apache.org/confluence/display/JMETER/TestRecording210">this wiki page before starting.</a> </subsection> <subsection name="&sect-num;.2 Basic Instructions" anchor="basic-instructions"> <ol> <li>Go to <code>JMETER_HOME/bin</code> and start JMeter with <code>jmeterw.bat</code> on Windows and <code>jmeter.sh </code> on Linux/Unix</li> <li>Select <code>Templates…</code> on the menu bar</li> <figure image="Select-Templates-Icon.png"></figure> <li>Select <code>Recording template</code> on the list</li> <figure image="Select-Recording-Template.png"></figure> <li>A complete <code>Test Plan</code> is generated</li> <figure image="Test_Generated.png"></figure> <li>In the <code>HTTP Request Defaults</code> element: <dl> <dt><code>Server name or IP</code></dt><dd>enter <code>example.com</code></dd> <dt><code>Path</code></dt><dd>leave blank</dd> </dl> <figure image="http-config/http-request-defaults.png"></figure> </li> <li>Return to <code>HTTP(S) Test Script Recorder</code>, and click the <code>Start</code> button at the top.</li> <figure image="Proxy_Run.png"></figure> </ol> This will start the JMeter proxy server which is used to intercept the browser requests. A file called <code>ApacheJMeterTemporaryRootCA.crt</code> will be generated in <code>JMETER_HOME/bin</code> folder. Install this certificate in your browser, if you don't know how to do it, read <a href="component_reference.html#HTTP%28S%29_Test_Script_Recorder">Installing the JMeter CA certificate for HTTPS recording</a> </subsection> <subsection name="&sect-num;.3 Configure your browser to use the JMeter Proxy" anchor="configure-browser"> At this point, JMeter's proxy is running. For this exercise, we will use Iceweasel/Firefox to view some pages on the JMeter website. <ol> <li>Start Iceweasel/Firefox, but do not close JMeter.</li> <li>From the tool bar, click <menuchoice> <guimenuitem>Edit</guimenuitem> <guimenuitem>Preferences</guimenuitem> </menuchoice> (or <menuchoice> <guimenuitem>Tools</guimenuitem> <guimenuitem>Preferences</guimenuitem> </menuchoice> or type <code>about:preferences#advanced</code> as URL). This should bring up the options. <figure width="914" height="566" image="firefox-network-settings.png"></figure> </li> <li>Select the <code>Advanced</code> tab, and <code>Network</code> tab</li> <li>Click <code>Settings</code> button near the top.</li> <li>On the new pop-up, check <code>Manual proxy configuration</code>. The address and port fields should be enabled now. <dl> <dt><code>Address</code></dt><dd>enter <code>localhost</code> or the IP address of your system</dd> <dt><code>Port</code></dt><dd>enter <code>8888</code>.</dd> </dl> <figure width="532" height="240" image="firefox-configure-proxy.png"></figure> </li> <li>Check <code>Use this proxy server for all protocols</code></li> <li>Click <code>OK</code> button. This should return you to the browser</li> </ol> </subsection> <subsection name="&sect-num;.4 Record your navigation" anchor="navigation-recording"> <ol> <li>With your browser, in the <code>Address</code> bar at the top, enter <code>http://example.com/index.html</code> (replace <code>example.com</code> with your websites address). and hit the <code>enter</code> key.</li> <li>Click on a few links on your sites pages.</li> <li>Close your browser and bring up the JMeter window.</li> </ol> Expand the <code>Thread Group</code> and there should be several samplers. At this point, the test plan can be saved as is. <note>If you forget to add default HTTP Request settings, you will have to manually delete the server name, and port.</note> <figure width="258" height="194" image="example-recording.png"></figure> In this sample, there aren't any default request parameters. If a particular request parameter is required by all pages, the request defaults is where one would add the entries. <ol> <li>Select <code>Thread Group</code> and change a few defaults: <dl> <dt><code>Number of Threads (users)</code></dt><dd>enter <code>5</code></dd> <dt><code>Ramp-Up Period (in seconds)</code></dt><dd>do not change</dd> <dt><code>Loop Count</code></dt><dd>enter <code>100</code></dd> </dl> <figure width="357" height="288" image="example-thread-group.png"></figure> </li> </ol> </subsection> <subsection name="&sect-num;.5 Validate the script" anchor="validate-script"> Now we need to validate the script before to run our test plan. Save the test plan. Right click on the <code>Thread Group</code> <menuchoice> <guimenuitem>Validate</guimenuitem> </menuchoice> <figure image="Validate-Test-Plan.png"></figure> Check with <code>View Results Tree</code> element if all is ok. </subsection> <subsection name="&sect-num;.6 Variabilize and Correlate the script" anchor="correlate-start"> In some scripts, we will need to: <ul> <li>Variabilize some input (login, password, search words, …)</li> <li>Correlate some data (session variable, …) between two requests</li> </ul> To variabilize, we can use: <ul> <li><code>CSV Data Set Config</code> to get input data from csv file</li> <li>JMeter functions like <code>__counter</code>, <code>__time</code>, …</li> <li>etc.</li> </ul> To correlate, we can get data from a request with <code>Post Processors</code> like <code>JSON Extractor</code>, <code>Regular Expression Extractor</code>, … and inject it in another request. To find data to correlate, the easiest way to do it is to use the Search function in <code>View Results Tree</code>. <figure image="Search-Correlation.png"></figure> </subsection> <subsection name="&sect-num;.7 Start the test" anchor="test-start"> At this point, we are ready to run our test plan and see what happens. When you're ready to run the test, there are two ways: <ol> <li>With the gui, but it's not recommended to big load test.</li> <li>With the command line.</li> </ol> Solution 1, with the gui, but just during debug phase, use CLI mode (Non GUI) for your load test. <menuchoice> <guimenuitem>Run</guimenuitem> <guimenuitem>Start</guimenuitem> </menuchoice> or use the keyboard and press <keycombo><keysym>Ctrl</keysym><keysym>R</keysym></keycombo> Before you start the test, add a <code>Summary Report</code> element and select it. As the test runs, the statistics will change until the test is done. At the end of the test, the summary report should look like this. <figure width="984" height="220" image="example-summary-report.png"></figure> While the test is running, in the upper right-hand corner, there should be a green circle. When the test is done, the circle should be grey. <figure width="132" height="39" image="example-running.png"></figure> Solution 2, in command line, use <code>jmeter -n -t [jmx file] -l [results file] -e -o [Path to output folder]</code> At the end of the test, an HTML report will be generated and available in [Path to output folder] used in command line. </subsection> </section> </body> </document>

xdocs/usermanual/jmeter_proxy_step_by_step.xml (189 lines of code) (raw):