Jose salt01

Notas del editor

• #8: Current and future international projects in science require the use of increasing computing and storage resources to handle and process larger volumes of information by researchers distributed across the world. The evolution in hardware ( CPU computing power, storage capacity, network bandwidth ) will not be able to cope with these requirements, that are significant not only in terms of resources but also on the organization for resource sharing. The term e-Science has been coined to refer to “large scale science that will be carried out through distributed global collaborations enabled by Internet”. One of the research priorities in FP6, 1.1.2.i, “Applied IST research addressing major societal and economical challenges”, includes “Computing and Information GRIDs” as a research topic for “Complex problem solving in science”. GRIDs can be defined as an advanced computational concept on which next generation Internet applications will be based: an architecture proposing a global solution to computational challenges and a framework for e-Science. Ian Foster and Carl Kesselman, developers of the Globus GRID middleware, define its possibilities for enabling Virtual Organizations: ”An infrastructure that enables flexible, secure, coordinated resource sharing among dynamic collections of individuals, institutions and resources”. Different research GRID projects are addressing different issues of this framework, from fabric management and basic middleware to portals and development of application kits. Also the real implementation in a wide European scale, using existing and new computing facilities, so implying research on key practical issues concerning installation and operation management, is being studied in testbeds in current EU projects like EU DataGrid, and will be the objective of new Integrated Projects in the 6th Framework Programme. Also RTD projects are addressing the use of the GRID technology on different application fields, from biology to astrophysics. The proposal pursuits one of the most ambitious objectives in GRID computing: the possibility of implementing a GRID facility for real-time applications. It should be clear that the force driving this network will come from the application side: as indicated before, e-Science poses such challenges that simply we will need Real Time GRIDs to solve them. In this sense, it is a further step along the path open by projects addressing interactive applications in the GRID framework, like the EU CrossGrid project. However it will require a non trivial advance on many technical issues, in particular those related to the support of QoS and the dynamical management of resources, and also a conceptual effort to produce an adequate design for operation in the framework of an Open Grid Service Architecture, where this meta-computing challenge is incorporating ideas from the Semantic Web approach. A mixture of research institutions, universities, and technology companies in the network will assure that the deployed RTGRID is of general interest and flexible enough to accommodate new interested participants, but at the same time solves real complex problems using in a coordinated way existing computing facilities, producing a clear synergy. The proposers do have extensive experience on distributed computing and in particular in GRID computing for interactive applications, including relevant roles in EU projects (CrossGrid, IST-2001-3224). The participants expect also to take part in a separate Integrated Project enabling the basic GRID infrastructure. Existing facilities, that will be integrated in the RTGRID, will range from mid to large size clusters locally managed, with heterogeneous hardware and software. One of the main points will be the collaboration with the corresponding National Network Providers, basic to progress on QoS implementation issues.