News

	Round Up from Barcelona
	News and views on ESFRI
	Congratulations to the winners!
	Bob’s thoughts on EGEE’09 Barcelona
	Looking back at EGEE’09: What was important for your activity?
	EGI Updates
	Hot off the presses from Barcelona
	You say e-infrastructure...
	The past and future of grids in four questions?
	Collaborating project sessions
	Go North!
	EGEE Road Show

Round Up from Barcelona

.

A reminder of the quick passing of time, EGEE’09 came and went in a flash. In late September, 21-25, 631 delegates from 43 countries gathered in Barcelona to contribute (and bid farewell to) the last of the Enabling Grids for E-sciencE conferences.

The week was exceptional for its rich scientific and technical presentations: attendees chose from 108 programme sessions and were shown 57 posters and 22 demonstrations. Delegates came from a wider-than-usual selection of the grid-interested community: notably representatives from 11 proposed European Strategy Forum on Research Infrastructures (ESFRI) projects, and from US and EC e-Infrastructure funding offices. It was also exceptional for the meetings of many boards and subgroups – particularly the European Grid Initative (EGI) Council and the meetings of proposed Specialised Support Centres (SSCs). Several delegates mentioned they had never been so busy during a conference week before. What happened during EGEE’09? Read on . . .

[ To top ]

News and views on ESFRI

EGEE’09 represented a seminal milestone in the preparations of EGI-linked projects and their services to user communities, as well as the planned research infrastructures projects known collectively as ‘ESFRI’. These are ambitious, large projects (similar to the Large Hadron Collider at CERN in scope) which will require mammoth computing support as they start up over the next five to ten years. Having a generic infrastructure – such as the one that can be coordinated through EGI – will help those projects build their distributed research environment – and avoid wasting time and resources in creating unique infrastructures.

On Tuesday John Wood, the chair of the European Research Area Board, opened the plenary session “Operational Grid Infrastructures after EGEE.” Wood highlighted key issues for European research in the future and shared his thoughts on developing a methodology for linking research infrastructures (i.e. the proposed ESFRI projects) and e-Infrastructures such as GEANT, DEISA/PRACE and EGEE/EGI.

“My own personal opinion is that the e-Infrastructure is the essential lifeblood of all infrastructures,” said Wood after the closing of the morning session. “Therefore to get a common approach – as much as possible – is vital.”

Wood believes some of the largest issues facing large scale research are: proper management of big projects, energy requirements, data issues (ownership, preservation, protection), national cooperation vs. national protectionism and training new researchers. See slides from his talk or read the full GridCast interview.

In current news, the first ESFRI project to be funded in FP7 (the current funding package from the EC), the European X-ray Free Electron Laser (XFEL), to be located at DESY in Hamburg, Germany, was signed in Berlin during the conference on 23 September.

Read about the signing from the European EFEL news page.

ESFRI and EGEE/EGI – what next?

A number of common organisational and technical needs emerged from the ESFRI presentations. [See Bob Jones’ slides from the final ESFRI session.] These projects may benefit by using common approaches to these, in collaboration with Europe’s pre-existing e-Infrastructures.

EGEE will be following up on the discussions and contacts made with ESFRI projects. In Brussels later this month, October, an EU Concertation meeting in Brussels will allow these discussions to progress, person-to-person contacts have been set between members of EGEE (within the Project Management Board) and representatives of the ESFRI projects. Also the recently formed European E-Infrastructure Forum may be a useful interaction point between ESFRI projects and e-Infrastructures.  

[ To top ]

Congratulations to the winners!

As at each EGEE event, conference attendees selected the best poster and demo by vote – in what is essentially the EGEE equivalent to the Oscar’s. In a particularly close competition, EGEE’09 delegates elected “Towards a reference model for the LifeWatch ICT infrastructure” by Vera Hernandez and Axel Poigne, both of the Fraunhofer Institute IAIS, as the best poster. (Download the PDF (2MB).)

The demo competition was particularly interesting this year with all demonstrators invited to submit a short video introducing their demo – you can watch these on YouTube at http://www.youtube.com/enablinggrids. David Manset (MAAT GKnowledge) and Giovanni Frisoni (IRCCS Fatebenefratelli) were honoured with the winning demo award for “neuGRID – A grid-brained infrastructure to understand and defeat brain diseases.”

[ To top ]

Bob’s thoughts on EGEE’09 Barcelona

It was a pleasure to be at EGEE’09 – I really felt an upbeat, positive atmosphere permeated the week. Looking towards the next few months I think we all have much more confidence in indeed ‘realising EGI’.

We managed to bring together all key stakeholders in e-Infrastructures, and our discussions were very productive. It was also excellent to see so many ESFRI projects present – promising rich interactions with e-Infrastructures for the future.

We tried a few new routes for increasing the conference buzz and sharing news and information in the run-up to and during EGEE’09. I thought the online presence worked really well and hope you took advantage of the videoed demos online, the conference blog on GridCast.org and the EGEE Twitter and Flickr feeds.

Though we were all busy during EGEE’09, now is the time to maintain that level of energy as we continue to hammer out the details of the future infrastructure and its satellite projects. Can we make a success of this next phase? More than ever the answer is a resounding ‘YES WE CAN!’

[ To top ]

Reflecting on EGEE’09:
What was important for your activity?

Cal Loomis, applications activity manager (NA4): “I think that the most important development at EGEE'09 was the planning for the evolution of the scientific clusters into SSCs.

“There were meetings for each of the scientific clusters (both for the scientific program and for the SSC evolution). This culminated with a lively session on the borders between the SSCs and various EGI-related projects. Although productive, this showed that there is lots of work to do to get to a viable situation.”

Florida Estrella, Collaborating Projects Liaison Office: “EGEE'09 showed that the links within the community of EGEE's collaborating projects remain extremely healthy as evidenced by the many sessions convened, booths hosted and demos and posters presented by these projects and by their contributions to the publication Success and Sustainability: EGEE Collaborating Projects' Achievements in 2009 and Future Plans.”

Catherine Gater, dissemination, outreach & communication activity manager (NA2): “I think our dissemination channels – such as the GridCast blog, Twitter and Flickr feeds – really enabled attendees to communicate with each other during the event, which I was very pleased to see.”

Robin McConnell, training activity manger (NA3): “The future of grid-training was our major topic at EGEE’09. During one of the sessions we discussed the readiness of various training groups for beyond the end of EGEE, when training will be an NGI task. We were also able to gather delegates opinions on future training needs.”

Christos Kanellopoulos, activity manager for international cooperation & policy (NA5): “EGEE’09 was the right venue for dialogue between ICT based infrastructures and projects implementing the European Strategy Forum on Research Infrastructures Roadmap. A common recurring theme coming from most of the ESFRI projects was the adherence to standards.

“EGEE has long invested in the development and adoption of standards and has been a key contributor in a number of standardization working groups within the Open Grid Forum. The importance of standards was stressed even further at the dedicated session ‘Standards and Certifications – Improving the quality of e-Infrastructure software’ that was held on the second day of the conference.”

Maite Barroso Lopez, grid operations, support & management activity manager (SA1): “I think EGEE09 has helped us to better understand the EGI/NGI landscape for grid operations and to facilitate the transition to it.

“Some highlights were the progress made on MPI deployment and validation tests, the introduction to the next generation of operation tools being deployed right now, and the evolution of the user support model.”

Xavier Jeannin, activity manager of networking support (SA2): “In the context of a multi-domain e-Infrastructures, close collaboration is needed between grid sites and NRENs to address monitoring, troubleshooting, advanced network services and network operations. The EGEE / TERENA joint workshop during EGEE’09 highlighted typical use cases of this collaboration and renewed the interest in a stronger cooperation between NRENs and Grids.”

Oliver Keeble, integration, testing & certification activity manager (SA3): “The conference provided a valuable opportunity for activity members to get up to date on the latest plans for middleware management in the future, and to provide their feedback face to face. This was an important milestone in the orientation of EGEE-III towards EGI.”

Francesco Giacomini, middleware engineering activity manager (JRA1): “EGEE'09 was the perfect occasion to meet people from different domains and discuss all together how to move forward, notably how to continue the evolution of the Grid middleware after the end of the project.”

[ To top ]

EGI Updates

One of the key news items of the week was the selection of an EGI Council chairman. During Thursday’s meeting the council elected Per Öster of CSC, Finland. In this position Öster will drive the agenda of the EGI activity and liaise with the proposal writing team in preparation for submitting the EGI proposal to the European Commission on the 24th of November.

Additionally, during the conference week the EGI Council allocated most of the global operational and infrastructure focused tasks to National Grid Initiatives. These allocations are now being integrated into the proposal. The remaining global tasks, primarily those related to middleware, are now available for all NGIs and EIROs (the EIRO forum is a collaboration of seven European Intergovernmental scientific Research Organisations, responsible for infrastructures and laboratories) to bid for. Allocations will be proposed in advance of the next EGI Council meeting in Brussels on 23 October.

[ To top ]

Hot off the presses from Barcelona

To highlight the innovative research accomplished with EGEE’s production grid, the EGEE press office sent out several releases during the week of EGEE’09. Particularly of note, Global grids tackle global science and Clouds predicted in Barcelona (covering the merging grids and clouds) were picked up by several news outlets. Download the full text of all releases below.

Global grids tackle global science
24 September 2009, Barcelona

Clouds predicted in Barcelona for annual grid conference
23 September 2009, Barcelona

Smart Technologies for Business and Science
23 September 2009, Barcelona

Grids and Health e-Science: Shaping the future of healthcare
21 September 2009, Barcelona

[ To top ]

You say e-infrastructure, I say cyber infrastructure... let’s collaborate: European Commission and National Science Foundation Sessions at EGEE’09

– From Catherine Gater, EGEE dissemination

At EGEE, the project office is a paragon of US-EC collaboration –European and American staff work literally side by side, often with much amusement generated on all sides by the quirks of each other’s cultures (the mysteries of London’s Cockney rhyming slang being the most recent source of interest/confusion). So I was keen to attend Tuesday’s session on EC and US work programmes, which particularly focused on opportunities for closer collaboration.

Presenting for the EC was Kyriakos Baxevanidis, Deputy Head of Unit for the GÉANT & e-Infrastructure Unit; from across the pond was Dr. José Muñoz, Director (Acting) of the Office of Cyber Infrastructure.

Baxevanidis started out by describing the EC’s motivations for funding e-infrastructure, including solving the major global challenges facing all of us of today, such as climate change, poverty, disease and finding new sources of energy. The EC also aims to accelerate globalisation, promote democratic values and ICTs worldwide.

[Read Catherine’s full account on the GridCast blog.]

[ To top ]

The past and future of grids in four questions?

Attendees of EGEE’09 formed a unique crowd. Many delegates had been part of the grid community before there was any sort of grid at all: when it was still a mere twinkle in the eyes of computer scientists. These people have seen this idea grow in to a reality. Did it happen as expected? What shape will the future take?

EGEE’s dissemination team took advantage of this grid “School of Athens” that gathered for the week to ask them four questions:

1. What do you think EGEE's main achievement has been?
2. What were its main challenges?
3. Did grids develop the way you expected?
4. If you had a magic wand, where would grids be in 5-10 years?

Unfortunately, we did not get to speak to everyone we would have liked. Would you like to share your perspective? Please write project-eu-egee-dissemination@cern.ch.

Cal Loomis, EGEE applications, CNRS:
(1.) The main achievement I think, by and far, is the production grid infrastructure that is used all over Europe by many users – I think that is clearly the best achievement of EGEE itself.
(2.) The main challenges were getting people together and to cooperate. I think the sociological and political problems of doing that are much harder than any of the technical ones.
(3.) Yes, I think grids developed in a direction I thought they would, albeit in a much slower than I expected. I expected they would come along technolocially much faster and I thought we would have a much bigger grid with more usage earlier – like just after European DataGrid. But I think more or less they've developed along the lines of what was foreseen in the beginning.
(4.) I think grids will still be there – I hope they will still be there – but I think they will look quite a bit different. What we are already discussing at this and other conferences is how to incorporate new technologies and new ways of using computing resources in the grid. And I think two areas where things will look very different is at the infrastructure level where I think we will see a lot of virtualization and cloud-linked technologies and at the user level where we will see things like Hadoop and other service platforms and application platforms to simplify particular types of calculations.

John Gordon, STFC-Rutherford Appleton Laboratory
(1.) The main achievement of EGEE has been to pull together all these countries so that you can easily and seamlessly submit jobs across national lines. That's been demonstrated by way some of the LHC (Large Hadron Collider) work has been running by countries that didn't expect to support the LHC.
(2.) The main challenges were meeting the conflicting requirements of different research areas.
(3.) I think yes, but more slowly than I expected.
(4.) I hope they will have integrated in some of the big European facilities like ILL - Institut Laue-Langevin, ESRF - European Synchrotron Radiation Facility and ITER - the International Thermonuclear Experimental Reactor. If they are integrated in to facilities like that, and aiding the scientists using them, then I think that will be a big achievement.

Stefan Lüders, CERN IT security officer
(1.) Well the first question isn’t easy for me to answer . . .
(2.) . . .but, on the second one, since I'm coming from security area, I think security is still a challenge. There is a cacophony of certificates, for example, when users sign on to the grid – I think this is still a challenge which needs to be overcome.
(3.) Grids didn't develop the way I expected. From the software applications I think it is much too heterogeneous. For example, there are many small sites which cannot just take something out of a box and become a Tier-1, Tier-2 or Tier-3 site. [Note: Visit here for an explanation of the WLCG Tier system.) This is what I find a little bit of a pity. I'm not a grid guy – but this is how I was expecting the grid to work: a university with a decent computer center who wanted to join the grid could take the free software, deploy it and then you become a regional center. But apparently it doesn't seem to be like this. I've seen too many parallel developments of the same thing and I think this is a waste of resources. I see the motivation behind it! But I think it is unfortunate.
(4.) For the future? I would like to see the grid really being what it is suppose to be: a computing service for research. One where people can hook up, submit their jobs, do calculations, universities can contribute in terms of computing power and it is free for everyone.

David Fergusson, National e-Science Centre, Edinburgh
(1.) The main achievement is creating an organisation that allows the support and development of a pan-national infrastructure.
(2.) It's biggest challenge was to bringing together people who all have their own visions and giving them a general vision that they could work to.
(3.) Did grids turn out as I thought they would? No.
(4.) In the future, I'd like to see grids supporting as wide a range of researchers as possible in a variety of fields, particularly in very large fields – such as across the life sciences, etc. So ideally, what we'd like to see at the EGI conferences, instead of having several hundred people, we'd see several thousand.

David Wallom, Oxford e-Research Centre
(1.) EGEE's main achievements have been to continue holding very disparate community together to build an infrastructure that is now in production.
(2.) Main challenge? Disparate forces trying – with vested self-interest – to maintain individual research communities and create their own infrastructures.
(3.) Grids have unfortunately not developed the way I thought they would. I was really hoping grids would become 'real infrastructure' supporting all research, connecting all communities together. We still haven't achieved that yet.
(4.) Grids or e-Infrastructure will become the underpinning on which all research will be done, whether or not it is high-throughput computing, high performance computing, remote instrumentation, visualization, all of these things should be available quickly and easily to the researcher communities.

Xavier Jeannin, CNRS, EGEE networking support
(1.) I think the main achievement for EGEE is that network support is operational, it allows the transfer of data across all of EGEE and that it is working now.
(2.) The biggest challenges, I think, are the multi-domain problems. The biggest challenge in the future is for the NRENS will be able to collaborate with each other in order to provide new advanced services and support for network operation for the grid.
(3.) Unfortunately no, in terms of networks, the situation has not evolved a lot.
The services (advanced network services and support for network operation for the grid) that were provided by the NRENS has been poor up till now, with no new services coming up.
(4.) I would like to see the multi-domain problems solved.

Fabrizio Gagliardi, Microsoft Research
(1.) The greatest achievement of EGEE is that it became real – out of so many initiatives in the early 90s, only EGEE has grown to production quality with a user community. It is real.
(2.) Biggest challenge? To keep a real user community, like physicists, happy, while at the same time being a generic multi-disciplinary infrastructure.
(3.) Did they develop as I expected? No – I think the dream of the early pioneers was to implement a pervasive, worldwide, unique computing infrastructure. That was simply too difficult for the technology we had available.
(4.) Where do I think grids will be? Disappeared! Like any computer technology they rely on the components, business model, environmental constraints, which they are built upon. In the future I think we'll see full virtualization of computing resources. The Grid has fulfilled its historical cycle. As it is today, it is not offering the best model for science and business in general, but this is with exceptions. This may be one of the last grid parties we will have. In the future we'll see virtualization of all computer services.

Markus Schulz, EGEE
(1.) The biggest achievements? Linking more than 250 computer centers from 5 nodes to several thousand has created a community that shares not only resources, but also the craft and art to operate large scale computer centers. All centers provide now services at a higher quality than a few years ago. Cooperation of centers has also massively improved the handling of security incidents.
One could say the flavor of middleware isn't important, common operations practices and tools make the difference.
(2.) What problems did we face? From the start on I had the impression that the technical problems were the easiest to solve. By linking many centers you not only link different computers, but also different grown cultures of operating resources and relating to users. Many site administrators viewed the grid as something that limits the level of control that they have over their center. To build the necessary trust into the project and the other sites was the biggest problem.
(3.) At the beginning we didn't understand the nature of the project. As a consequence we tried to use management structures and procedures absolutely adequate for a local project, but quite difficult for a project with subtasks that have 20 partners with some only providing a fraction of a person.
Instead to provide building blocks to the users for grid computing we developed too many end to end solutions. As a consequence, the basic functionality that is used at massive scale isn't always as efficient and reliable as it could have been. The user's extensions to these services luckily compensate this to a large extent.
(4.) Where will grid computing be in the future? A misquote of Chairman Mao might describe the future best: "Let a thousand flowers bloom". Grid computing is on one hand moving to better established standards, on the other hand it becomes an integrating tool for other paradigms of distributed computing.
Already now people have build prototypes to link technologies such as Cloud Computing, voluntary computing (BOINC) with production grids. In addition more user communities will be using grids without being aware of it, I guess portals to grids will have a great future.
However, I have been notoriously bad in predicting the future of computing, despite being aware of emerging technologies early on. ( My bank statement is a clear proof of this )
If I had a magic wand to influence the future? Tempting! there are a few collaborating partners that are sometimes very strong headed, maybe there the wand could come handy. I guess some would use it on me......
Otherwise the lesson learned is to evolve step by step based on the experience gained. There is no shortcut for experience when moving new concepts into reality, especially on a massive scale. For the rest slightly more generous funding to keep the infrastructure going and work on improved middleware is all that is needed.

Highlights from collaborating project sessions

From Federico Ruggieri, Session Regional Extensions of Grid Infrastructures
This session showed the great interest existing players have in stepping forward in terms of coordination through the CHAIN project. A largely agreed topic was the need to better understand and investigate the concept of 'virtual research communities' and the role they should have in the present and future infrastructure projects.

From Ake Edlund and Ruben Montero, Session Grids and Cloud Computing: Perspectives and Early Experiences
Grid-meets-Cloud, a very timely and interesting topic, was addressed by the session Grids and Cloud Computing: Perspectives and Early Experiences convened by the BalticGrid-II and RESERVOIR projects. In-depth experiments, hands-on installations and first-hand experiences with clouds and virtualisation techniques were presented, followed by a panel discussion on the complementarities and differences of these concepts.

From Ignacio Martin Llorente, Session EGEE and RESERVOIR projects
An “interest group” will form from partners collaborating in the preparation of the Memorandum of Understanding between RESERVOIR and EGEE, and other partners working on similar approaches to bring cloud to grid computing. The group aims to define an architecture for the virtualization of grid sites (local cloud deployments) that could be enhanced to provide infrastructure services to remote users (public cloud deployment), proposing a “cloud solution” for EGEE and EGI.

[ To top ]

Go North!

Come April we’ll be heading north to Uppsala for the final User Forum which will be held in collaboration with EGI and the Nordic DataGrid Facility in Uppsala, Sweden, 12-16 April 2010.

With the establishment of sustainable European and National e-Infrastructures this event will be the ideal place for European e-Infrastructure users to meet, share experiences, and shape the future of European e-Infrastructures. More updates in the coming months, see you there!

[ To top ]

The EGEE Road Show

– From Neasan O’Neill, EGEE’s Press and Events Manager

The annual EGEE conference always marks the beginning of the busy conference season for the project.

Next up e-Challenges in Istanbul. EGEE will share a booth with EGI and GridTalk at the meeting to give the grid infrastructure and its success stories a greater visibility. Now in it's 19th year the conference is supported by the European Commission and attracts people from research, industry and government to discuss and disseminate knowledge and best practice related to ICT. The aim for this year's meeting is to improve technology transfer from the European Research Area to industry.

In November the collaboration heads to one of the world’s premier computing events, SC09 in Portland, Oregon. With people from every area of computer research, as well as representatives from the media and governmental bodies, this is a showcase of the future of computing, in research industry and the home. Once again the EGEE presence is coordinated with EGI to showcase the past and current success of the grid but also the planned transition to a sustainable European grid infrastructure.

The final two conferences of the year are UK e-Science All Hands Meeting and the 5th IEEE International Conference on e-Science both being co-located in Oxford in the UK. With representatives from all areas of research the all hands meeting is one of the largest eScience conferences in the UK and a chance to engage new researchers from throughout Britain and disseminate the benefits of grid computing to them. The IEEE meeting is designed to bring together international and interdisciplinary research communities, developers, and users of e-Science applications.

eChallenges, Istanbul - 21 to 23 October
SC09, Portland Oregon - 14 to 20 November
UK eScience All Hands Meeting, Oxford UK - 7 to 9 December
5th IEEE International Conference on e-Science, Oxford UK - 9 to 11 December

[ To top ]