The three levels of the software testing process are unit testing, integration testing, and system testing.
Unit testing tests the isolated functionality of each individual unit.
Integration testing tests the functionality of combining units to ensure they work together as intended.
System testing tests the software as a whole.
In my reading about integration testing, I noticed that at some point during this testing phase the system should be tested as a whole. Is system testing just a redundant and more in depth testing methodology focused on this? Sean Coughlin answers this question in his blog “Understanding System Testing in Software Engineering” (https://blog.seancoughlin.me/understanding-system-testing-in-software-engineering) by providing a clear understanding of what system testing is and what it tests for.
What is System Testing?
System testing is a software testing phase that assesses a fully integrated software system’s compliance with requirements, covering both functional and non-functional aspects. These non-functional aspects are what makes system testing so comprehensive compared to integration testing when testing the software as a whole. Examples of non-functional testing include; performance testing, security testing, and usability testing.
Coughlin uses an e-commerce website as an example to illustrate system testing and each of its components. Functional testing ensures each feature of the website works as intended, Performance testing ensures the website still works even during peak user times. Security testing ensures the website is robust against cyber threats and that data is handled correctly. Usability testing ensures the website is user-friendly across multiple platforms.
Testing frameworks exist that allow for automated test such as simulating heavy loads for performance testing. User testing on the other hand is typically done manually as it focuses on a user’s experience. This means having a range of people, from developers to stakeholders, testing end-to-end use of the software.
Overall system testing is testing an application for real world use. This goes much deeper than just functionality of the software. This is the last phase of testing before the software is released to the final product for acceptance testing. Meaning it is important that developers properly system test to identify and fix issues during this phase.
The scope for system testing is even larger than I could have imagined. Coughlin states, “I like to think of system testing as more of an umbrella term that captures lots of different testing forms.” which perfectly describes how system testing encapsulates testing all aspects of a software.
CS Learning 