Abstract:

Choosing the right design, architecture and technologies for an application can be difficult for architects and engineers, especially when dealing with big and fast data. Indeed, these platforms need to deal with the data growth and variety, while generating insights in a timely fashion. These constraints, combined with the wide range of available storage and processing technologies, are
hard to satisfy. On the other hand, and with the lack of clear methodologies, best practices and validation tools, testers struggle to perform end-to-end evaluations for the developed Big Data solutions. The need for a generic benchmarking solution, able to integrate with any multi-layered Big Data architecture, is manifest. In this paper, we present and evaluate BABEL, a generic, scalable, distributed and end-to-end Benchmarking Platform for Big Data architectures