In this presentation, we will cover a variety of research subjects, ranging from Cloud Computing to Big Data applications. We will first focus on providing a quick overview about Cloud Computing and Big Data, necessary to understand the open issues that will be presented in the rest of the seminar.
The remainder of the presentation is organised around four themes. With a bottom-up presentation style, we will first talk about the impact of virtualisation on the performance of data-intensive applications, from the I/O point of view. Then, we will discuss about our current research effort in the design of efficient scheduling mechanisms for the allocation of resources in MapReduce clusters. Building on the knowledge of what are MapReduce jobs, the presentation will continue on high-level languages to express parallel programs and optimisation thereof. Finally, we will cover the application layer and illustrate a number of use cases (i.e. algorithm design) that our Research Group covers in cooperation with industrial partners. The research work outlined above is funded by a recent project, called BigFoot: for additional information, please visit http://www.bigfootproject.eu
CSS e script comuni siti DOL - frase 9957