ACADEMIA

• The analysis of the results of CERN experiments has been made possible by development of a new technology called grid that has allowed the creation of a distributed supercomputing infrastructure that currently has over 300,000 distributed computing cores in 300 institutions in over 40 countries.
• The grid plays a key role in enabling the analysis of the results of the ATLAS and CMS experiments at CERN.
• CESGA is responsible, along with other institutions, to ensure the availability of grid infrastructure used to perform simulations.
• CESGA develops and manages websites and accounting metrics to monitor the largest distributed supercomputing infrastructure in the world, EGI, providing thousands the supercomputing capacity needed to address grand challenges such as Higgs boson search.
• The experience of CESGA grid infrastructure is supported by its participation in the international Grid initiatives, since 2001 in the EGEE, EELA, Crossgrid, Int.eu.grid, EGI and EMI.
• CESGA participates in the development and verification of software and grid middleware using experiments at CERN.

According to officials at CERN, the grid is key to the success of the ATLAS and CMS experiments, which have contributed more directly to the discovery of new particle. This was highlighted in the presentation of preliminary results that have revealed the observation of an unknown particle, possibly the Higgs boson. The WLCG, Worldwide LHC Computing Grid, the Grid infrastructure created to distribute, store and analyze data from the LHC experiments at CERN, uses software developed by the Centro de Supercomputación de Galicia, CESGA, through different European grid initiatives.

Behind the CERN particle accelerator there is a collaboration involving more than 8,000 scientists from 347 institutions from over 40 countries working in different Virtual Organizations (VOs). Processing data relating to the ATLAS VO (one of the experiments responsible for the detection of the Higgs boson) requires 10,234,519,228 hours supercomputing on the European grid infrastructure that is part CESGA. The second experiment responsible for the discovery, the CMS VO processed data that have required 3,766,770,932 computing hours using the same infrastructure.

CESGA experience in the grid comes from its participation in European grid initiatives, developed and tested software for EGEE (Enabling Grid for E-Science) now a CGA (European Grid Initiative) to EELA (E-science grid facility for Europe and Latin America), EMI (European Middleware Initiative), and int.eu.grid Crossgrid.

CESGA is also responsible for monitoring the IBERGRID Operations Center, together with the Institute of Physics of Cantabria, IFCA and the Physics Laboratory and Instrumentação Experimental Particle Lisbon, LIP. Its mission is to coordinate operations in the grid infrastructure in Spain and Portugal and monitor the correct operation of the centers Peninsula computing involved in LHC experiments at CERN.

Portals and Accounting Metrics

The proper development of the European grid supercomputing project requires constant management, not only with the infrastructure but also in human capital. Also, certain objectives are set for each project period in order to promote growth and development of the grid. To do this, we have identified a set of metrics to be collected by each task project regularly, both globally and for each federation or country.

CESGA has developed a Metrics Portal for EGI, which not only allows the collection of such metrics by national officials and the project, but also it is able to provide estimates for many of them. The site automatically collects information from various sources and generates the corresponding metric. It enables objective measurement of project progress and track their evolution in a simple and secure way and is based on updated data, enabling the generation of immediate reporting.

Since 2006, CESGA develops and maintains the Accounting Portal to display and analyze the consumption of processors, number of jobs and efficiency, among other measurements associated with use of supercomputers in the grid. It also allows you to generate usage statistics resources (for experiment, date, investigator, institution, etc. ..), and the issuance of reports, including those associated with use of multiple federation commitments to specific projects like ATLAS.

EMI Project

Additionally, CESGA through its participation in the EMI project (European Middleware Initiative), in collaboration with the IFCA and LIP, has performed the integration of EMI middleware with GridEngine, the queuing system used in many of Europe's supercomputers, including CESGA's own. GridEngine has advanced distributed computing capabilities, very useful to the institutions participating in the grid, and now have the option of choosing this queue system as well as LSF Torque solutions and benefit from support and assistance provided by EMI for users of its middleware.

Data deluge

Until next December, they will present the definitive analysis of the ATLAS and CMS experiments, this grid will support traffic of more than one million daily computers working. The researchers have access to supercomputing resources needed to understand the flood of data from sensors installed in the Large Hadron Collider (LHC) that are used to conduct the experiments at CERN.

Resources available to the LHC SuperComputing

Moreover, as a member of EGI, CESGA is part of the LCG, which is the Tier 1 in Spain PIC (Port d'Informació Scientific) in Barcelona, and coordinates the participation of Tier 2 in seven other scientific institutions, including the Laboratory for High Energy University of Santiago de Compostela. Also, CESGA has supercomputing resources arrangement of the four major LHC experiments: ATLAS, CMS, ALICE and LCHb.

More information: Carlos Fernandez Sanchez, Dept. of Systems, CESGA carlosf@cesga.es 981569810 http://www.cesga.es