Energy consumption in big data environments – a systematic mapping study

Abstract

Big Data is a term that describes a large volume of structured and unstructured data. Big Data must be acquired, stored, analyzed and visualized by means of non-conventional methods requiring normally a big set of resources, which includes energy consumption. Although Big Data is not new as a phenomenom, its explosion of the interest in literature is recent and its study in new scenarios presents several gaps. On the other hand, Green IT is also a growing field in computing, given the increasing role of IT in energy consumption in the world. Green IT is aimed to reduce IT-related energy consumption and overall IT environmental impact. In order to investigate the reported initiatives regarding the Big Data and Green IT with a focus of energy consumption, the authors conducted a systematic mapping on the topic. The search strategy which was used resulted in 28 relevant studies which were relevant to the topic. We found that a majority of the studies performed present algorithms designed to reduce the energy consumption in data centres. The rest of the studies present benchmarks and energy measurements, reviews, proposals of hardware-based solutions, as well as studies which give an overview of one or more aspects on Big Data.