Large scale grids
The idea of this special issue on large scale grids came after a successful meeting that has been co-located with Europar'2005 in Lisboa. This meeting, namely Grid@Large 1 was organized by INRIA 2 , the University of Nice (France) and CoorGrid 3 . I take the opportunity to thank Luc Bouge ´and Igor Rosenberg for the setting and organization of the meeting and for asking me to join the team.
The workshop was made up with the contribution of a dozen of invited speakers. It gave ''updated visions of large-scale deployment, user management, financial issues, and security in grids''. It was also a matter of concerns on ''grid achievements, large-scale grids in other domains, physical maintenance, failure recovery, networking, software/hardware upgrade''.
Among the rich conclusion of the workshop, I would like to mention the observation of participants saying that ''a deployment phase of information technologies is starting, but the application interaction stage remains complicated''. The lessons learned during the workshop pointed out the difficulties in/with Performance (while structure can grow and spread, local memory does not hence programming difficulties), Simplicity (grids should be easy to use for people who are already coming from other sciences than Computer Science or Engineering), Interoperability (the data format and communication models should follow the user needs), Upgrading (schemes where deprecated technology is working with state-of-the-art implementations), Security/ Reliability (do we leave it to the user?) and Management (distributed know-how can be hard to handle).
To carry on with such topics and to check the reality of such assertions, we decided to implement this special issue and to give the word to researchers rather than experts. The aim was to find good projects with meaningful results that demonstrate the utility, the relevance of grids, but large scale grids. Many discussions with Denis Trystram (Parallel Computing Representative) have been necessary to converge towards an adequate format.
Then a call for papers was setup. The special issue aims at gathering papers in the fields of Grid, Distributed Systems, Networking. The special issue was opened for contributions in connected fields: P2P systems, Sensors Networks, Networking, Large Scale Heterogeneous Distributed databases. This way, we expected to be able to confront concepts of these various areas and then, investigate which ones may present mutual interest for the different communities. The ''traditional'' Grid community has up to now developed environments based on 'in-house' ideas and skills (Globus, Unicore, etc.), with little attention to sibling communities. The challenges set by ''large-scale'' use of Grids will necessarily demand importing ideas from other technologies, in the first place peer-to-peer approaches, and component-based software engineering as mentioned previously.
Research in large scale distributed systems requires a new methodology since most of the past work have not investigated the impact of large scale notions on algorithms and applications. So, new tools, benchmarks and use cases have to be investigated.