While designing a performance test, it is extremely important to understand which of the all business scenarios the test should include. While doing so, the load subjected on and the criticality of each those scenarios need to be considered. This is not an easy job and exhaustive discussion need to happen between the business teams, the project managers and the technical teams. The exhaustive discussions and analytics data obtained from web analytics tools can lead to identification of the right scenarios to be tested. However the job doesn’t end there. Identifying the load distribution across those scenarios during the test is also a key task and needs to be performed astutely. In this blog, let’s understand how to distribute the load across different business scenarios. Following are the steps as are performed by the business teams –
Connect with the sales & marketing teams and understand what sells in the market. Understand the features from them as they visualize seeing in the product. Also understand the nature of other products in the market from the competitor and what performance features the competitors provide to the end users. This would help in setting the business target.
Web log information for applications which are already live:
Download the weblog data and from the same extract following key information –
For new websites:
Your business insights:
Based on user activity data, web analytics data and your business insights, try to understand the load that can be expected on different modules (components) of the application and distribute the load accordingly.
After collating this information and finalizing the same, following are some key activities that the teams must complete so as to be compliant with the process (an overall idea presented here)
Thus, these are some ideas that can be used to distribute load across key performance scenarios. It is however essential to understand that activities of end users and the load patterns may change as the product evolves and as the project changes and hence it is important to perform the above activity (of load distribution) in a repeated manner after a specific period of time so that any change in load distribution can be appropriately studied and implemented in the project. This is how one can target to take his test more close to production scenario and thus be able to perform the test as user perform their actions in production.