Wednesday May 9, 2018
Many blogs have discussed about functional and non-functional testing, but only a few have differentiated between the two. So, before we discuss the differences between the two, it is important to first take a brief look at functional and non-functional testing.
Functional testing, as the name suggests, revolves around the functional aspects of the system or subsystems and how efficiently the system executes its functions. This testing is carried out by specifications provided by the client, or use cases provided by design team. Prior to conducting non-functional testing, functional testing should be carried out.
So What About Non-Functional Testing?
Non-functional testing is done against the non-functional requirements, or which are not covered in functional testing. Non-functional requirements tend to be those that reflect the productivity, quality, stress of the system, specifically from the user’s perspective.
For example, functional testing might reveal the test for inputting data and its expected behaviour, whereas non-functional testing reveals a time taken for a document to get saved. Therefore, non-functional testing demonstrates how the system behaves as compared to what the system is expected.
Functional Testing Includes The Following Testing Techniques:
White Box Testing: Includes testing the internal structure of the system code
Black Box Testing: Based on the test objective, derive test cases
Unit Testing: Identify bugs and errors right from the start of the development life cycle.
Smoke and Sanity Testing: Compare the system with specifications to make sure its functioning accordingly
Integration Testing: Test individual software components to verify interaction between various interfaces
Non-Functional Testing Involves Following Sub-Types of Testing’:
Compatibility Testing: Performed to ensure compatibility of a system, application, or website
Configuration Testing: Testing an application with multiple combinations of software and hardware to find out the optimal configurations
Load Testing: Determine a system’s behaviour under both normal and anticipated peak load conditions.
Performance Testing: Evaluating of the quality or capability of a product.
Recovery Testing: Determine how quickly the system can recover after it has gone through system crash or hardware failure
Security Testing: Determine if an information system protects data and maintains functionality
Scalability Testing: Measure its capability to scale up or scale out in terms of any of its non-functional capability.
Stress Testing: Stretches software to its limits and checks its durability.
The strategies of testing differ for both in functional, as well as, non-functional testing. For functional testing, the test strategies are attained using automated and manual tools. The widely used functional testing method includes black box testing, where the tester isn’t required to test the internal source code, but validates the system using valid and invalid inputs.
All functional tests like, API testing, unit testing, Boundary value analysis etc., have a specific output that has an expected output. After entering valid data, the system should work as intended. Functional tests are scripted with pass and fail criteria.
Coming to the non-functional testing, the tests conducted under this require more creativity and technical expertise compared to the functional ones. This is because you are developing a product keeping the end user workflow in mind, rather than what input gives what output.
While some non-functional tests can be defined in fail and pass criteria, many are more of an objective and require opinion of the testing team. Therefore non-functional tests are written in measurable and quantifiable way wherever needed. Having a good communication and requirement gathering from the client is essential when you perform non-functional testing
The ultimate conclusion here is, having knowledge about the differences these 2 testing types carry can help you during your test planning, so that you are clear about what you are going to test. Nevertheless, you system should pass both the test before its even released to the final end user. Failing to release a fully-fledged product that meets the needs of the client’s demands can damage your company’s reputation and reduce overall product sales.