You are hereTest Construction & Execution

Test Construction & Execution


Benefits

  • No wasting of time associated with poor testing practices resulting in tests which have to rebuilt to answer requirements effectively.
  • Elimination of political issues associated with "blaming the tool" or "blaming the test" when the results point to a performance issue
  • A measure of risk in deployment based upon both the test results as well as the differences between production and test.

Overview

Testing is about the solid application of the scientific method. Every performance test conducted by our staff adheres to the standards set forth by Aristotle over two thousand years ago. Unfortunately few organizations in the performance test industry adhere to such well established, rigorous and defensible test practices. As a result tests are often of low quality, are poor predictors of what will happen in production and often provide misinformation about the scalability of newly deployed applications.

Does each one of your internal tests take into account best practices noted below for execution? Send us one of your LoadRunner scenario definitions and we'll examine your test matched to industry best practices.

We object to the relatively poor standard in which tests are conducted across the profession and follow the principals of the scientific method. Every test conducted by our organization contains

  • A confidence factor as a predictor of performance in production taken from the differences between the production and test environments
  • A set of working hypotheses derived from your expressed requirements
  • Solid, hardened steps to reproduce which are incorporated into the scripts representing your business processes
  • Expected results for each step, which are examined by each business process script
  • A control factor, by which it may be determined whether the test itself is functioning correctly and that the results obtained are valid for comparison to your requirements.
  • Documented initial conditions/state which allows the test to be returned to the identical state for another test
  • Reproducibility. As a function of the initial conditions and the control factors each of our tests will produce a substantially similar result when executed a second, third, fourth or more times assuming that no undocumented, uncontrolled elements are present in the test bed
  • High defensibility. Because we are adhering to scientific method principals set down centuries ago and refined since we can defend each test against accusations of errors due to the test definition or testing method.