INFN Grid

INFN Grid
Name	INFN Grid
Established	2002
Type	Distributed computing infrastructure
Location	Italy

INFN Grid is a national distributed computing infrastructure developed by the Istituto Nazionale di Fisica Nucleare to support large-scale scientific computing for high-energy physics, astrophysics, and computational science. It federates computing centers, storage systems, and network resources across Italian research institutions to provide resources for experiments such as those at CERN, collaborations with European Grid Infrastructure, and projects linked to ESFRI initiatives. The effort bridges resources from universities, national laboratories, and international consortia including ties to INFN Sezione di Bologna, INFN Sezione di Napoli, and other INFN sections.

History

The genesis of the project is rooted in early European grid efforts inspired by initiatives at CERN and the Particle Physics and Astronomy Research Council collaborations that followed the success of the Large Hadron Collider computing model. In the 2000s INFN aligned with projects such as LCG and EGEE while coordinating with national partners like Istituto Nazionale di Statistica and research centers such as CINECA and ENEA. Subsequent phases saw integration with grid middleware from Globus Toolkit and adoption of services interoperable with European Grid Infrastructure and international experiments including ATLAS (experiment), CMS, and LIGO Scientific Collaboration. INFN nodes participated in distributed computing readiness tests for Worldwide LHC Computing Grid and contributed to federated resource sharing with institutions like University of Bologna, Sapienza University of Rome, and Politecnico di Milano.

Architecture and Infrastructure

The architecture federates Tiered computing centers inspired by the Tier architecture of Worldwide LHC Computing Grid and leverages storage and compute farms at sites such as INFN Laboratori Nazionali di Frascati and INFN Laboratori Nazionali del Gran Sasso. Network interconnectivity relies on links to backbone providers including GARR and interoperation with GEANT for pan-European connectivity. Compute resources include CPU farms, batch systems like HTCondor, virtualization layers influenced by OpenStack, and storage systems interoperable with dCache and StoRM. The infrastructure integrates data management patterns from collaborations like ALICE and uses resource catalogs analogous to those developed for EGI.

Services and Middleware

Core services include job submission, data transfer, workload management, and information systems built upon middleware components from gLite predecessors, ARC, and Globus Toolkit derivatives. Data transfer services exploit protocols used by GridFTP and storage services reference implementations like dCache and XRootD. Middleware supports user communities from Astroparticle Physics experiments through interfaces compatible with tools developed for Ensemble Antares and clients used by BaBar (experiment), BESIII, and KM3NeT. Identity and authorization tools integrate with systems akin to VOMS and job orchestration borrows concepts from HTCondor and PBS Professional workflows employed by institutions such as University of Padua.

Projects and Collaborations

INFN Grid participates in European projects including EGEE, EGI, and collaborations with CERN experiments like ATLAS (experiment) and CMS, as well as astrophysics consortia including VIRGO and LIGO. It has engaged with computational initiatives at CINECA and contributed to infrastructure efforts under frameworks such as Horizon 2020. The grid has supported experiments at Gran Sasso National Laboratory and projects in computational biology and climate science collaborating with institutions like ENEA and Istituto Nazionale di Geofisica e Vulcanologia.

Operations and Resource Management

Operations are coordinated through regional resource centers and operational centers modeled after the operational procedures of Worldwide LHC Computing Grid and European Grid Infrastructure. Resource allocation uses accounting and monitoring tools comparable to GridICE and service monitoring approaches from Nagios and Zabbix installations in academic centers such as University of Turin and University of Pisa. Service-level coordination draws on practices from EGI Operations Centre and regional operations teams liaise with experiment-specific computing coordinators from ATLAS (experiment), CMS, and ALICE.

Security and Authentication

Security architecture uses X.509 certificate mechanisms and federated authorization systems akin to VOMS and integrates with identity providers from research infrastructures such as PRACE and GEANT. Incident response and vulnerability management follow operational patterns seen in CERT-EU and national CERT teams collaborating with centers like INFN Sezione di Roma. Secure data handling for experiments such as BABAR, LHCb, and KM3NeT incorporates encryption for data transfers via protocols like GridFTP and access controls modeled on ESO and CERN Computer Security practices.

Impact and Scientific Contributions

INFN Grid enabled data processing and analysis for major discoveries and ongoing research in particle physics, gravitational-wave astronomy, and astroparticle studies, supporting collaborations such as ATLAS (experiment), CMS, VIRGO, and LIGO. It fostered advancements in distributed computing techniques later adopted by cloud projects like OpenStack deployments at research centers including CINECA and informed European e-infrastructure policies through participation in European Grid Infrastructure governance. The infrastructure underpinned publications from University of Bologna, Sapienza University of Rome, and Politecnico di Milano groups and contributed to training programs connected with Erasmus Mundus and national doctoral schools.

Category:Distributed computing Category:Istituto Nazionale di Fisica Nucleare