ATLAS Off-Grid sites (Tier-3) monitoring

ATLAS is a particle physics experiment on Large Hadron Collider at CERN. The experiment produces petabytes of data every year. The ATLAS Computing model embraces the Grid paradigm and originally included three levels of computing centers to be able to operate such large volume of data. The ATLAS Distributed Computing activities concentrated so far in the “central” part of the computing system of the experiment, namely the first 3 tiers (CERN Tier-0, the 10 Tier-1s centers and about 50 Tier-2s). This is a coherent system to perform data processing and management on a global scale and host (re)processing, simulation activities down to group and user analysis. With the formation of small computing centers, usually based at universities, the model was expanded to include them as Tier-3 sites. Tier-3 centers consist of non-pledged resources mostly dedicated for the data analysis by the geographically close or local scientific groups. The experiment supplies all necessary software to operate typical Grid-site, but Tier-3 sites do not support Grid services of the experiment or support them partially. Tier-3 centers comprise a range of architectures and many do not possess Grid middleware. Thus, monitoring of storage and analysis software used on Tier-2 sites becomes unavailable for Tier-3 site system administrator and Tier-3 sites activity information becomes unavailable for virtual organization of the experiment.