Next Generation Grid(s)

Executive Summary - The expert group convened in two meetings at Brussels to provide an expert's view on the 5 to 7 year research priorities for the European context. This group's opinion was that current Grid implementations lack many essential capabilities, which would enable the vision of complete resource virtualization. Three perspectives are defined to envision Next Generation Grid(s) (NGG): the end-user vision in which the simplicity of the Grid is exemplified; the architectural vision of a large-scale, selforganizing network; and the software vision of a programmable system. Each of these visions is described in detail to support the assignment of priorities to the research agenda. In order to realize those visions, a number of research priorities have been identified. They fall into three categories, the first one considering the properties of a Next Generation Grid, the second one specifying the facilities provided by an NGG, and the third one comprising the models that are needed for orchestrating Grid services. For clarity, a number of research topics have explicitly been excluded as falling outside the scope of this paper.