To enable predictive and scalable performance, a testing strategy should be injected throughout the development lifecycle.
During development, developers should leverage profilers and load-testing tools, such as the ones tested here, to see which portions of a site or application are taking a long time to execute. These areas can then be addressed, and additional testing can be executed to validate the results.
In the staging process, new rounds of profiling and load testing should be executed to revalidate the results achieved during development. A set of baseline tests, which exercise the site or application as it will be used under expected production workload, is a must. Production usage should be simulated as closely as possible through the use of virtual users and variable-based input.
Beyond baseline load testing, additional rounds of tests should be executed during staging to evaluate scalability and performance given expected business growth over time. These types of tests can proactively help forecast when additional IT infrastructure will be needed to support the business. When what you are testing has been deployed, you'll likely want to use only passive monitoring tools to measure how well a site or application is performing and to keep overhead under control.