Grid to cloud migration of scientific applications, using dynamically created cloud clusters

Abstract

The paper presents an experimental study on the migration of scientific applications from Grid to Cloud Cluster infrastructure. The research aims to highlight the main conceptual, technical, and technological issues of such a process by exploring solutions for distributed processing, execution and management of a complex task, increased data management, and gathering of processed data. As a case study, the Grid based gSWAT application is used, supporting the SWAT hydrological model calibration over the Grid infrastructure. The main challenge of the research is the mapping of the gSWAT platform and data model onto the services and resources provided by the infrastructure, platform, and software Cloud levels. The Cloud related experiments are based on the OpenNebula and Hadoop technologies.