Modeling Expands Value of Performance Testing for Big Data Applications

Abstract

Performance testing of Big Data applications is performed typically on small test environment with limited volume of data. The results of these types of tests do not take into consideration differences between test and production hardware and software environment and contention for resources with many applications in production environments. In this paper we will review application of the modeling for extending the results of performance testing, predicting how new application will perform in production environment. We will review how modeling results can be used to evaluate different options and justify decisions during design, development, implementation and performance management of the production environment.