Proper launch of the EGI - A challenge for the European Grid Community
While grids are commonly understood as distributed computing infrastructures, they are primary about the collaboration. It is therefore not surprising that the first international pusher for the large scale grid infrastructure was a community behind the high energy physics scientific experiments. A development that initially started as a support activity of one scientific community became soon a movement that attracted scientists from many other disciplines and a recognition at the highest national political levels.
The initial efforts culminated in a series of the EGEE projects, run between 2004 and 2010 and supported by the European Commission within the 6th and 7th Framework Programmes. While successful, these projects also demonstrated the shortcomings of an infrastructure based on standard projects and taken care of by very heterogeneous group of supporting institutions, ranging from university departments to proper infrastructure building bodies.
The community, with the support from the commission, started in 2007 a design study project EGI_DS—European Grid Initiative Design Study—to properly define the conditions and requirements for a large scale trans-European grid infrastructure. The project, coordinated by CESNET and led by the author of this article, produced a series of documents, with the EGI Blueprint and accompanying deliverables on functional requirements setting the framework for the future grid infrastructure in Europe. The proposal prepared by the EGI_DS was inspired by the successful history of trans-national computer networks and their organizational model, based on a combination of a national and international coordination. The basic building blocks for the future European Grid Infrastructure were therefore defined as National Grid Initiatives (NGI), organizations that are responsible for the national coordination of grid infrastructure and that are also representing individual countries at the international level. The NGIs are complemented by the EGI.eu, an organization founded as part of the EGI_DS activities, with headquarters in Amsterdam in The Netherlands. EGI.eu together with the NGIs form a basic foundation of the European grid collaboration and are responsible for the coordinated operation of the European Grid Infrastructure. The governance is guaranteed through the EGI Council, where each participating NGI, as well as international body like CERN, does have its representative.
The long term sustainability of this model should be provided through national funding of NGIs that understand benefits of close international collaboration and coordination of activities. The first step towards this model, the so called Transition towards the EGI, is financially supported by European Commission through its moderate co-funding (at the level of one third of total expenses) for a four year EGI InSPIRE project. The project started on 1st May 2010 and its success or failure will determine not only the validity of the EGI model, but the fate of the grid EGI collaboration itself. With NGIs and not individual research institutes its primary partners, the infrastructure managed by the EGI InSPIRE is much more application neutral than its EGEE predecessor. While the project itself includes a very basic funding for the heavy user communities like the HEP, Bioinformatics or Computational Chemistry, the major challenge lies in providing an infrastructure that will be attractive for scientists irrespective of their organizational and scientific affiliation. A close collaboration with user communities—most notably those behind large ESFRI projects—is essential, but EGI must not abandon also smaller and more dispersed scientific communities. This task is even more challenging as the major projects targeting these communities were not accepted by the Commission and EGI InSPIRE must look for new partners and new ways of collaboration with user communities. One way to do this is through emphasizing the roots—grids as a collaboration platform, connecting scientists across Europe. Properly managed infrastructure open to any scientist pave the way towards trans-European grid infrastructure that will be widely accepted as the basic eInfrastructure of the future, with no fear for its sustainability. Let’s support the EGI InSPIRE project to properly respond to this challenge.
Ludek Matyska, Institute of Computer Science Masaryk University, Czech Republic