Deployment of job priority mechanisms in the Italian Cloud of the ATLAS experiment

An optimized use of the Grid computing resources in the ATLAS experiment requires the enforcement of a mechanism of job priorities and of resource sharing among the different activities inside the ATLAS VO. This mechanism has been implemented through the VOViews publication in the information system and the fair share implementation per UNIX group in the batch system. The VOView concept consists of publishing resource information, such as running and waiting jobs, as a function of VO groups and roles. The ATLAS Italian Cloud is composed of the CNAF Tier1 and Roma Tier2, with farms based on the LSF batch system, and the Tier2s of Frascati, Milano and Napoli based on PBS/Torque. In this paper we describe how test and deployment of the job priorities has been performed in the cloud, where the VOMS-based regional group /atlas/it has been created. We show that the VOViews are published and correctly managed by the WMS and that the resources allocated to generic VO users, users with production role and users of the /atlas/it group correspond to the defined share. © 2010 IOP Publishing Ltd.