Cloud Elasticity for Implementing an Agricultural Weather Service

In this paper, we describe a weather service dedicated to farming called CAP2020. The service involves interpolation operations accessing massive data. We explain the centralized implementation of this application and show the limits of this architecture. Then we present a Cloud-enabled distributed solution ensuring load balancing and data replication to take advantage of the elastic capabilities of Cloud solutions. Distributed implementation provides better quality of service and scalability for CAP2020 application in anticipation of increasing use and/or application data, but also to adapt to punctual and seasonal farming needs.