Performance testing basics in Testing Fundamentals - Build an Automation Script

Measure website response time under load

Preconditions (2)

Step 1: Open the performance testing tool

Step 2: Create a test plan with 50 virtual users

Step 3: Set the test to send requests to the website homepage URL

Step 4: Run the test for 5 minutes

Step 5: Monitor the response times and error rates during the test

✅ Expected Result: The average response time should be under 2 seconds and error rate should be 0%

Automation Requirements - Apache JMeter

Assertions Needed:

Verify average response time is less than 2000 milliseconds

Verify error rate is zero

Best Practices:

Use realistic number of virtual users to simulate load

Add assertions to check response time and error rate

Use timers to simulate real user think time

Run tests for sufficient duration to get stable results

Automated Solution

Testing Fundamentals

import org.apache.jmeter.config.Arguments;
import org.apache.jmeter.engine.StandardJMeterEngine;
import org.apache.jmeter.protocol.http.control.HeaderManager;
import org.apache.jmeter.protocol.http.control.gui.HttpTestSampleGui;
import org.apache.jmeter.protocol.http.sampler.HTTPSamplerProxy;
import org.apache.jmeter.reporters.ResultCollector;
import org.apache.jmeter.samplers.SampleSaveConfiguration;
import org.apache.jmeter.testelement.TestPlan;
import org.apache.jmeter.threads.ThreadGroup;
import org.apache.jmeter.threads.gui.ThreadGroupGui;
import org.apache.jmeter.util.JMeterUtils;
import org.apache.jorphan.collections.HashTree;

public class PerformanceTest {
    public static void main(String[] args) throws Exception {
        // Set JMeter home for JMeterUtils
        String jmeterHome = System.getenv("JMETER_HOME");
        JMeterUtils.setJMeterHome(jmeterHome);
        JMeterUtils.loadJMeterProperties(jmeterHome + "/bin/jmeter.properties");
        JMeterUtils.initLocale();

        StandardJMeterEngine jmeter = new StandardJMeterEngine();

        // Create Test Plan
        TestPlan testPlan = new TestPlan("Performance Test Plan");

        // Create HTTP Sampler
        HTTPSamplerProxy httpSampler = new HTTPSamplerProxy();
        httpSampler.setDomain("example.com");
        httpSampler.setPort(80);
        httpSampler.setPath("/");
        httpSampler.setMethod("GET");
        httpSampler.setName("Homepage Request");
        httpSampler.setProperty(TestElement.TEST_CLASS, HTTPSamplerProxy.class.getName());
        httpSampler.setProperty(TestElement.GUI_CLASS, HttpTestSampleGui.class.getName());

        // Create Thread Group
        ThreadGroup threadGroup = new ThreadGroup();
        threadGroup.setName("Load Test Users");
        threadGroup.setNumThreads(50);
        threadGroup.setRampUp(10);
        threadGroup.setDuration(300); // 5 minutes
        threadGroup.setScheduler(true);
        threadGroup.setProperty(TestElement.TEST_CLASS, ThreadGroup.class.getName());
        threadGroup.setProperty(TestElement.GUI_CLASS, ThreadGroupGui.class.getName());

        // Create Test Plan tree
        HashTree testPlanTree = new HashTree();
        HashTree threadGroupHashTree = testPlanTree.add(testPlan, threadGroup);
        threadGroupHashTree.add(httpSampler);

        // Add Result Collector
        SampleSaveConfiguration saveConfig = new SampleSaveConfiguration();
        saveConfig.setTime(true);
        saveConfig.setLatency(true);
        saveConfig.setSuccess(true);
        saveConfig.setResponseCode(true);
        saveConfig.setResponseMessage(true);
        saveConfig.setThreadName(true);
        saveConfig.setDataType(true);
        saveConfig.setEncoding(true);
        saveConfig.setAssertions(true);
        saveConfig.setSubresults(true);
        saveConfig.setResponseData(false);
        saveConfig.setSamplerData(false);
        saveConfig.setAsXml(false);

        ResultCollector logger = new ResultCollector(saveConfig);
        logger.setFilename("performance_test_results.jtl");
        testPlanTree.add(testPlanTree.getArray()[0], logger);

        // Run Test Plan
        jmeter.configure(testPlanTree);
        jmeter.run();

        // After test, parse results file to assert average response time and error rate
        // (Parsing and assertions would be done in a separate step or tool)
    }
}

This code sets up a JMeter test plan programmatically.

First, it configures JMeter environment and loads properties.

Then it creates a test plan with a thread group of 50 users ramping up over 10 seconds and running for 5 minutes.

An HTTP sampler sends GET requests to the homepage of example.com.

A result collector saves test results to a file.

The test plan tree is built and the test is run.

Assertions on average response time and error rate should be done by analyzing the saved results file after the test run.

Common Mistakes - 4 Pitfalls

Using too few virtual users that do not simulate real load

Not adding assertions to check response time and errors

Running tests for too short a time

Hardcoding URLs without using variables

Bonus Challenge

Now add data-driven testing with 3 different URLs to test homepage, login page, and search page

Show Hint

Practice

1. What is the main goal of performance testing in software development?

easy

A. To find spelling mistakes in the user interface

B. To check how fast and stable the software works under heavy use

C. To test if the software installs correctly

D. To check if the software looks good on different screens

Solution

1. Step 1: Understand the purpose of performance testing

   Performance testing measures speed and stability when many users use the software at once.

2. Step 2: Compare options with this purpose

   Only To check how fast and stable the software works under heavy use talks about speed and stability under heavy use, matching the goal.

3. Final Answer:

   To check how fast and stable the software works under heavy use -> Option B

4. Quick Check:

   Performance testing = speed and stability check [OK]

Hint: Performance testing checks speed and stability under load [OK]

Common Mistakes:

• Confusing performance testing with UI or installation testing
• Thinking performance testing checks visual design
• Mixing performance testing with functional testing

2. Which of the following is a common tool used for performance testing?

easy

A. Photoshop

B. WordPress

C. Excel

D. JMeter

Solution

1. Step 1: Identify tools related to performance testing

   JMeter is a popular open-source tool designed for performance testing.

2. Step 2: Eliminate unrelated tools

   Photoshop is for images, Excel for spreadsheets, WordPress for websites, none for performance testing.

3. Final Answer:

   JMeter -> Option D

4. Quick Check:

   Performance tool = JMeter [OK]

Hint: Pick tools known for load testing like JMeter [OK]

Common Mistakes:

• Choosing unrelated software like Photoshop or Excel
• Confusing content management systems with testing tools
• Not knowing popular performance testing tools

3. Consider this scenario: A performance test runs 100 users simultaneously on a website. The average response time is 2 seconds, but 10 users experience 10 seconds delay. What does this indicate?

medium

A. The test is invalid because average time is low

B. The website handles load well with no issues

C. There is a performance bottleneck affecting some users

D. All users have the same experience

Solution

1. Step 1: Analyze the response times

   Most users get 2 seconds, but some have 10 seconds delay, showing inconsistency.

2. Step 2: Understand what inconsistent delays mean

   Long delays for some users suggest a bottleneck or problem under load.

3. Final Answer:

   There is a performance bottleneck affecting some users -> Option C

4. Quick Check:

   High delay for some = bottleneck [OK]

Hint: Look for slow responses in some users to spot bottlenecks [OK]

Common Mistakes:

• Ignoring slow user delays and trusting only average
• Assuming test is invalid without checking details
• Thinking all users have same experience

4. You wrote a performance test script but it always stops with an error after 5 users start. What is the most likely cause?

medium

A. The test script has a limit on concurrent users set too low

B. The website is too fast and closes the test

C. The test tool does not support more than 5 users ever

D. The internet connection is too slow

Solution

1. Step 1: Understand the error context

   The test stops after 5 users start, suggesting a limit reached in the script or tool settings.

2. Step 2: Evaluate options for cause

   The test script has a limit on concurrent users set too low explains a low user limit in the script, which matches the problem. Other options are unlikely causes.

3. Final Answer:

   The test script has a limit on concurrent users set too low -> Option A

4. Quick Check:

   Test stops early = user limit too low [OK]

Hint: Check user limits in script when tests stop early [OK]

Common Mistakes:

• Blaming website speed or internet without evidence
• Assuming tool cannot handle users without checking settings
• Ignoring script configuration for concurrency

5. You want to test a web app's performance under increasing load. Which approach best simulates real user behavior during the test?

hard

A. Gradually increase the number of virtual users over time while running typical user actions

B. Run 1000 users all clicking the same button at the exact same time once

C. Run a single user repeatedly refreshing the homepage for hours

D. Test only the database without simulating user actions

Solution

1. Step 1: Understand realistic load testing

   Real users increase gradually and perform varied actions, not all at once.

2. Step 2: Compare options to realistic behavior

   Gradually increase the number of virtual users over time while running typical user actions simulates gradual load increase with typical actions, matching real use best.

3. Final Answer:

   Gradually increase the number of virtual users over time while running typical user actions -> Option A

4. Quick Check:

   Realistic load = gradual increase + varied actions [OK]

Hint: Simulate real users by increasing load gradually [OK]

Common Mistakes:

• Testing all users clicking simultaneously unrealistically
• Using only one user repeatedly
• Ignoring user actions and testing only backend