From the first grid jobs to the EOSC

Arjen van Rijn reflects on the challenges ahead

This issue of the EGI newsletter is dedicated to 15 years of operations. The EGI federation is now an established computing platform for research, with demand growing at an average 20% per year.

These 15 years have been memorable, for me personally. In this period the successful BiG Grid project took place from 2007 to 2012: the roll-out of the Dutch distributed computing infrastructure for research. From 2013 the Big Grid infrastructure, which included an LHC Tier-1 facility, was adopted by SURF, the Dutch e-infrastructure organization. In 2014 Nikhef acquired co-funding for its share of the Dutch Tier-1 for the period until 2019. Meanwhile, Amsterdam Science Park was selected to host the EGI Foundation, established on 8 February 2010.

I was a member of the first EGI Executive Board until 2014. In 2012, I was appointed one of the Dutch e-IRG delegates and then I became involved in evaluation and monitoring of ESFRI projects and proposals. This episode has taught me much about the interplay between generic (horizontal) and disciplinary (vertical) infrastructures, both on the national and the European level.

And now we have the European Open Science Cloud. In its Staff Working Document (SWD), the European Commission has defined the EOSC as a virtual environment to offer free at the point of use, open and seamless services for storage, management, analysis and reuse of research data, across borders and scientific disciplines. The challenges for the EOSC reside on the interface between vertical and horizontal e-infrastructures. The latter have the potential of being efficient and effective, pooling hardware, software and, even more importantly, people and expertise together instead of building disciplinary pillars.

EGI has been experiencing how integration of computing with data and discipline-specific analytics services, is key for the successful adoption by users. The current use by research communities in physics, earth sciences, neuro-informatics, chemistry, biology and many more, are proof of this success.

According to the SWD, the EOSC will most likely be implemented as a federation of infrastructures. In my view, the generic part of the EOSC will be the federation of national e-infrastructures. e-IRG has been advocating governments and funding agencies to improve e-infrastructure coordination at the national level. These strong national e-infrastructure building blocks will enable frictionless cross-border research collaborations and contribute to European e-infrastructure landscape.

The EGI Federation has become a strong and visible part of this landscape. 23 NGIs and CERN are today contributing to build a federation spanning hundreds of datacenters, operating as a remarkably smooth whole, to serve research communities in Europe. A federation par excellence!

I expect that the EOSC will be the incentive for further consolidation in the European e-infrastructure ecosystem, extending well into its national foundations. This consolidation has also been mentioned by Andreas Veispak (DG CONNECT), in his talk at the recent EOSC-hub week in Prague.

Challenging times lie ahead to explore with our friends in this ecosystem, EUDAT, G√ČANT, OpenAIRE, PRACE, and our collective constituency, the best possible way forward to make and keep our researchers happy!

More information:

Arjen van Rijn is the Institute Manager of Nikhef and chairs the EGI Council since January 2019.

Issue 34

Table of contents

PDF

Subscribe to the EGI newsletter: