Science Gateways for the Broader Take-up of Distributed Computing Infrastructures

Abstract

Accessing Distributed Computing Infrastructures (DCI) such as resources of computational Grids has long been regarded as the privilege of the knowledgeable few. Early adoption of these technologies required significant patience due to the instability of the infrastructure and the everchanging nature of middleware and low level tools. Early adopters of DCIs also required low level technical knowledge and understanding of distributed computing paradigms. As a result the take up of Grid infrastructures has been relatively low. While some researchers published significant results using large amount of computational power, the majority of scientists found it too complex and too difficult to engage with. Although cloud computing technologies address some of these problems, the utilisation of vast amount of resources to solve computation and data intensive tasks still stayed well beyond the computational knowledge of a typical domain scientist, such us a biologist or a chemist. Science gateways offer the potential to significantly change this trend and open the utilisation of DCIs to a much larger audience. Low level command line interfaces will always be mas-