How to Perform Manual Performance Testing?

In this Educational Tutorial on Performance Testing, we have presented a comprehensive discourse on Mobile App Performance Testing.

The significance of a fast and receptive website is directly associated with drawing in more patrons and providing an enhanced user experience.

Performance testing is vital as it aids in ascertaining how an application behaves under varying levels of load and user interactions.

=> Click Here For the Whole Performance Testing Tutorials Series

However, can performance testing be manually executed as functional testing? In this article, we will delve into this facet of performance testing.

The timely delivery of a quality product with complete test coverage is crucial for any application.

Manual and automation testing have consistently been two concurrent strategies to aid testers in achieving this aim. Nonetheless, manual test execution is time-intensive, costly, and susceptible to mistakes. Automation testing resolves these issues, saving time and effort in the long term.

Here are some significant advantages of automation testing:

• The same test script can be employed for multiple executions by altering the test data when necessary.
• Compared to manual testing, automation testing dramatically lessens execution time.
• Automation testing provides reliable and consistent results.
• Automation testing covers the gaps where manual testing may not be effective (for example, testing web services).

Likewise, for performance testing, employing automated test scripts delivers more relevant and trustworthy results compared to manual test cases.

• Performance testing involves evaluating software applications’ performance and understanding their behaviour under specific conditions or expected workloads. Response time is a critical parameter in performance testing.
• The primary objectives of performance testing are to assess the speed, stability, and scalability of applications.
• Performance testing measures the efficiency and performance of computers, networks, software programs, or devices.

Here’s an overview of the performance testing process:

Both performance testing and automated testing involve conducting simple checks using various tools. Test automation typically combines functional testing, load testing, and performance testing.

When a functional test automation script is enhanced to measure the recording and rendering times of objects/screens, the script converts into a performance monitor.

The whole process goes as follows:

1. Initiate the load generator to test the server’s response time during peak hours or under load.
2. Once the targeted load is achieved, begin the functional test automation suite to measure the actual time taken to accomplish tasks in a fully rendered environment.
3. This process offers valuable insights into the user experience during high load or peak hours.

Recommended Reading: => Discover how to conduct load testing using HP LoadRunner

The subsequent diagram illustrates the blend of “Performance” and “Test Automation”:

Can performance testing be executed manually?

Now, let’s address the question, “Can manual testing alone, without the use of automated test scripts, carry out performance testing?“.

As stated previously, combining performance and automation testing yields reliable results. Nevertheless, there are circumstances where the performance of an application can be evaluated manually as well.

Let’s check out some examples:

Example 1: Assume a tester needs to swiftly assess the performance of a certain feature within a specific time frame. Here, a manual test can be conducted to determine the page load or response time.

Example 2: Consider a web application based on Java. To carry out manual performance testing in this scenario, multiple active sessions of the application need to remain open for testing. Manually, the tester can ascertain the number of active sessions, the number of open database connections, the count of running threads, and monitor CPU usage, memory consumption, and other relevant parameters.

Tivoli Performance Viewer (TPV) (Link) can be installed to monitor these parameters of interest. The test generally involves deploying the application on a server, accessing it from various client machines, and allowing multiple threads to run.

Here’s how Tivoli Performance Viewer is utilized to monitor the WebSphere Application Server (WAS)’s performance:

WebSphere Application Server (WAS) is a software framework/product that hosts Java-based web applications. The performance monitoring data obtained by using Performance Viewer for any WAS instance can be examined to recognize factors contributing to extended response times and implement necessary performance enhancements.

The subsequent procedure is followed to record performance using Tivoli Performance Viewer:

• Select the server to be monitored. For instance, “server1” is chosen in the diagram below.
• During the application’s runtime, server activity is observed by viewing real-time data gathered by Performance Monitoring Infrastructure (PMI) within the administrative console. This collected PMI data aids in performance optimization.

Here’s a diagram illustrating this process:

In this, the checkbox for “Server1” is selected, and on clicking “Start Monitoring,” the Performance Viewer commences the monitoring of the application server’s PMI data.

(Click on the image to magnify)

The subsequent diagram offers an overview of the settings and views available for “Server1” in Tivoli Performance Viewer’s left panel and the corresponding monitored data in the right panel.

(Click on the image to magnify)

• User configuration and settings for logging can be adjusted to view the server performance logs within the administrative console. These settings also impact the performance of the application server.
• The real-time data recorded by the Performance Viewer, such as server performance logs, can be reviewed later as well.

Example 3: Assessing an application’s performance for various network speeds.

Example 4: Conducting a quick initial performance test on a database to gather necessary information.

Therefore, in these scenarios, fundamental performance testing can be performed without solely depending on automated test scripts. Diagnostic or monitoring tools might be required, but total automation is not always essential.

When can manual performance tests be advantageous?

A non-performance tester can swiftly conduct manual tests without the need for setting up automated test suites or extensive knowledge of test automation.

This approach is particularly beneficial in the following circumstances:

• When performance testers require a speedy evaluation of a product’s response under specific conditions.
• When executing sanity performance tests while troubleshooting potential issues.

What to avoid?

While it’s possible for performance testers to simulate performance test conditions from multiple client machines, individually accessing the servers, this approach isn’t very effective. It’s essential to realize that not everything can be done manually.

Conclusion

Performance testing is perceived as one of the most intricate kinds of software testing, necessitating substantial technical knowledge. Automated test scripts bring considerable value to the testing effort, but searching for new techniques to combine them with manual testing can be useful. Manual testing can augment performance testing efforts.

Do you concur? What other modifications of performance testing have you delved into?

Stay tuned for our imminent tutorial on Website Performance Testing!

=> Click Here For the Entire Performance Testing Tutorials Series

PREVIOUS Tutorial | NEXT Tutorial