Every software application is different. As such, not all applications need to be performance tested. Additionally, the same application may not require repeated performance testing for every release/cycle. This begs the question, ‘under what conditions will an application need performance testing?’
Identifying risks to the performance of an application is a good way to start. You should look at risk factors that include the user population, application type, application technology and changes to the application features and function. Let’s look at each of these in detail.
1. User population
Individuals who use your application are crucial to helping you make the decision on its risk potential. This becomes even more important if your customer base is continually growing. You should ask the following questions:
- Who are the end users of my applications?
- How many concurrent users access my application?
- What is the volume and speed at which users grow each year?
- Are users purchasing products/services from your application?
- Are they external or internal users?
- What is the user spread across the geography?
- How much traffic do you anticipate during Black Friday, Cyber Monday and seasonal holiday sales on your website?
2. Application typeAs the application type can dominate the risk factors, the questions that you should ask include:
- Is the application an online retail website?
- Is the application integrated with other key applications within your enterprise architecture? If yes, this may require a performance test for every release.
- Does the application run several batch processes with a strict window of processing?
3. Application technology
Normally organisations do not make frequent technology platform changes from release to release. If your organisation is moving to new and future technologies such as mobile, cloud, IoT and big data, these application platforms must be performance tested.
If you are adding a new technology or replacing an existing one, it may pose a risk, which requires performance testing of the hardware change, alterations in the system architecture, network upgrade and/or application migration to a different data centre.
4. Application features and functions
The amount of code change or new code deployment in an application can create new application performance risks. It is critical to understand the impact of code changes and the effect on functionality. You should ask questions such as:
- What is the volume of code change and its impact on existing application features?
- Is there a change in nature of the data being processed in the application?
- Is there an introduction/change in web services?
5. Production issues
If performance issues were discovered once an application went into production, it would require a detailed analysis of server resource utilisation patterns, usage trends and error logs for further risk mitigation.
Carrying a performance risk assessment will help you answer the question, ‘under what conditions will an application need performance testing?’ Performance testing cannot be simply ignored, because once you understand it, you can assess risks, make informed decisions and plan for the future. Know more about Mastek’s Testing Services. Connect with us at firstname.lastname@example.org