Measuring your application for performance is something that is one of those things that sometimes gets pushed to the end of milestones. This means that when you do test for performance, if the application is not performing well due to any number of reasons, you have to go back and fix it – going back and fixing things is always costlier and riskier than doing them right in the first place.
At Nagarro, we tend to stay away from this model by training our team members to follow a philosophy that advocates giving importance to performance throughout the development lifecycle. As the diagram below shows, this includes measuring and testing for performance iteratively throughout the development process.
Why is this hard? At Nagarro, we have test benches setup along with a formal performance testing process that the project teams use to measure the performance of their applications. However, being ‘formal’ and how they are setup, it is difficult to use them multiple times during the project. Ideally, I would like performance to be measured in our projects on a daily basis.
We ought to be able to measure the performance of our applications on the fly (on demand).
