Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 76: Computing 1

T 76.3: Talk

Wednesday, February 29, 2012, 17:15–17:30, VG 0.111

Data access at Tier2s in ATLAS — Adil Aktas¹, Guenter Duckeck², Anton Gamel¹, Torsten Harenberg³, •Sergey Kalinin³, Gen Kawamura⁴, Kai Leffhalm⁵, Joerg Meyer⁶, Andreas Petzold⁷, Cedric Serfon², Jan-Erik Sundermann¹, and Rodney Walker² — ¹Uni Freiburg, Freiburg im Breisgau — ²LRZ/LMU, München — ³BUW, Wuppertal — ⁴Uni Mainz, Mainz — ⁵DESY-ZN, Zeuthen — ⁶Goegrid, Göttingen — ⁷KIT/SCC, Karlsruhe

The LHC experiments have developed very complex and powerful data distribution systems which provide detailed information on overall data distribution and usage. However, this covers the actual data usage only partially, since access from locally submitted jobs is not included. A set of tools has been developed at Bergische Universitaet Wuppertal and Ludwig-Maximilians-Universitaet Munich to fill this gap for sites which use dCache as the storage system. These tools also provide means to obtain filtered accounting information and reveal data access patterns which are not available in the standard dCache monitoring. We will present the typical usage patterns for two Tier2 sites (BUW, LRZ) for the case of ATLAS but also other VOs active at BUW.