EGI Workshop, Monday 22nd September
EGI will be a successor to current European grid infrastructure projects like EGEE and DEISA. Requires participation from exactly one national grid organisation in each country who wishes to participate. This seems challenging in some countries. Some have projects with no national authority and some have more than one competing body. Some representatives don’t seem that interested in an International infrastructure even though I know their country has active EGEE sites and projects.
The Blueprint published 10 days ago is a big improvement on the one from June. It now gives a better idea of why we need a central coordinating organisation for an EGI (EGI.org). Not everyone is happy with the size of it though (50FTEs).
The speaker from DEISA seemed to exist entirely in an HPC-only universe. I don’t think he understood the question about data from simulation needing to flow to non-HPC Data Repositories. Most national networks have enough capacity for these flows but need the protocols and services defined. I think there is also a use case for users at a university being able to send and receive their data from/to their local HPC centre via grid protocols.
The EGEE Panel Session was rather disappointing. It didn’t seem to address the key issues and didn’t give the audience enough time to participate. I suspect the organisers were afraid there would be no audience input. As if!