Performance testing is the practice of determining how applications, hardware, networks, databases, … will perform under a particular workload. The main goal is to improve speed, scalability, stability and reliability.

Performance issues have a nasty habit of turning up late in the application life cycle, the later you discover them, the greater the cost and effort to resolve (source O’reilly). In the figure below you see:  

• The ‘planned’ line where (performance) issues are tackled in the development phase and the application immediately starts to provide benefit to the business 
• The ‘actual’ line is what unfortunately often happens in businesses where development and deployment targets slip and performance issues need to be fixed in production.  

At Brightest we emphasize the need to: 

• Define clear goals: what do we want to do, what to measure, KPI’s, who are the stakeholders 
• Understanding the outcome of the performance test: how do we interpret terms like latency, throughput, percentiles, average times, …  
• Actions: based on the outcome, what are the steps to take to improve the scalability, reliability, resource usage, … 
• Benchmarks: when we run tests we also want to compare them to previous results, to industry bests and other companies. We consider quality, time and cost to better understand the outcome of the tests