Efficient Management of Geographically Distributed Big Data on Clouds

Nowadays cloud infrastructures allow storing and processing increasing amounts of scientific data. However, most of the existing large scale data management frameworks are based on the assumption that users deploy their data-intensive applications in single data center, few of them focus on the inter data centers data flows. Managing data across geographically distributed data centers is not trivial as it involves high and variable latencies among sites which come at a high monetary cost. In this report, we introduce an uniform data management system for disseminating scientific data across geographically distributed sites. Our solution is environment-aware, as it monitors and models the global cloud infrastructure, and offers predictable data handling performances for transfer cost and time. In terms of efficiency, it leverages for applications the possibility to set the tradeoff to be done between money and time and optimizes the transfer strategy accordingly. A prototype of our system has been implemented in the Windows Azure Cloud, and we obtain some encouraging results from the extensive evaluations.