The initial set of presentations reflected on the emergence of the EGI Design Study, its motivations and achievements over the last two years, and the work that has taken place over the last 6 months on the establishment of EGI.eu in Amsterdam.
The EGI model will be supported initially through EC projects. A summary of the EGI-InSPIRE, EMI, SAFE and ROSCOE projects were presented. One of the main outputs of the EGI_DS project were the Blueprint and Functions documents which identified the tasks that need to be supported within the EGI model. An analysis of the submitted proposals indicated that the vast majority of these tasks would be supported by one or more of the project proposals.
The first day finished with an analysis of the next steps being undertaken by the EGEE-III project to provide a smooth transition to the EGI model. These range from improving the overall quality of the infrastructure and providing VO specific availability and reliability views, to collaborating with other projects exploring new virtualisation technology. Establishing accredited trainers in all of the NGIs and closely aligning the training infrastructure with the production infrastructure remains a priority. These and other issues will be reviewed by EGEE-III now that the projects that EGEE will transition too have been submitted.
The final day provided a series of perspectives on EGI from outside the immediate EGI ecosystem - how would other infrastructures and communities interact with EGI from within Europe and outside of Europe? The perspective from collaborating infrastructure providers in Russia, Malaysia and the Open Science Grid were presented, followed by views from the Bioinformatics and HEP communities.
Finally, the closer interaction of European infrastructure providers were discussed. How will EGI (high throughput computing - and increasingly high performance computing) interact with PRACE (petascale computing) and the European networking infrastructure provided by GEANT? These discussions have already been taking place for several years now within the European E-Infrastructures Forum.
Closer collaboration between these infrastructures seems to be inevitable in the next round of EC projects. This will no doubt be welcomed by the user communities who use resources from multiple providers and would like to see improved integration between these resource providers. Currently, these users have to learn different tools and procedures to access these different resources.