Inspired by the 8th e-Infrastructure Concertation Meeting
Catherine Gater wraps up the event
More than 100 proposals received, 38 funded – this was the outcome of the e-Infrastructures Call 7 from last year, as presented by Enric Mitjana, from the European Commission, the coordinator of the call, at the 8th e-Infrastructure Concertation meeting at CERN on 4th and 5th November.
EGI-InSPIRE of course numbered among those projects, as did e-ScienceTalk, organiser of the event with the European Commission.
Speaking in the impressive venue of the Globe at CERN, Steven Newhouse, Director of EGI.eu, gave a overview of the six major distributed computing infrastructures that have kicked off in the last few months: EGI-InSPIRE, European Middleware Initiative (EMI), Venus-C, StratusLab, European Desktop Grid Initiative and the Initiative for Globus in Europe (IGE).
These projects can all contribute components to the recently launched flagship initiatives, The Digital Agenda and Innovation Union, part of the objectives for Europe 2020, the strategy for European growth over the next decade. The e-Infrastructure projects will also play a key role in sheltering us from the data deluge predicted to hit us in the coming years.
Working together with coordinating standards project SIENA, the six projects have identified some of the real barriers to broader adoption of e-Infrastructures. In the past, the sustainability of e-infrastructures has not been clear, a problem if you are a researcher worried about your data long term. This is now partially addressed by the new EGI structure using sustainability models adopted in other communities.
Until recently, advances in ICT have kept up with increasing needs for data analysis, but this is changing with the expected wave of data from the ESFRI projects. The fact remains that computing beyond your desktop is difficult for many – but virtualisation and new computing models should help simplify things.
Another issue is that current services do not always match the needs of new users and a flexible, federated production infrastructure will be better placed to meet these. With the advent of cloud computing, it is also important to compare up front the value, benefits and costs of public and commercial offerings.
So this is all food for thought for the projects over the coming years – but while cogitating on these barriers, they also realised some of the unique selling points that should not be forgotten. Often the compute resources are near the storage resources, co-locating processing near the data, all of which linked by high-speed networks. Moving large quantities of data around commercially can be time-consuming and expensive, so this is an advantage. For e-Infrastructures, high speed networks are ‘free’ at the point of use.
The projects also have more than ten years of trust built up in the community – sharing resources in a trusted environment is now second nature. In that time, the community has also established close links to its data-oriented users and has developed a strong understanding of the patterns in the ways that researchers use data, standing them in good stead for the challenges ahead.
