CERN Accelerating science

ATLAS Slides
Report number	ATL-SOFT-SLIDE-2016-012
Title	Scaling up ATLAS Event Service to production levels on opportunistic computing platforms
Author(s)	Benjamin, Douglas (Duke University, Department of Physics) ; Caballero, Jose (Brookhaven National Laboratory (BNL)) ; Ernst, Michael (Brookhaven National Laboratory (BNL)) ; Guan, Wen (Department of Physics, University of Wisconsin) ; Hover, John (Brookhaven National Laboratory (BNL)) ; Lesny, David (University of Illinois at Urbana-Champaign) ; Maeno, Tadashi (Brookhaven National Laboratory (BNL)) ; Nilsson, Paul (Brookhaven National Laboratory (BNL)) ; Tsulaia, Vakhtang (Lawrence Berkeley National Laboratory and University of California, Berkeley) ; van Gemmeren, Peter (Argonne National Laboratory) ; Vaniachine, Alexandre (Argonne National Laboratory) ; Wang, Fuquan (Department of Physics, University of Wisconsin) ; Wenaus, Torre (Brookhaven National Laboratory (BNL))
Corporate author(s)	The ATLAS collaboration
Submitted to	17th International Workshop on Advanced Computing and Analysis Techniques in Physics Research, Valparaíso, Valparaíso, Chile, 18 - 22 Jan 2016
Submitted by	[email protected] on 12 Jan 2016
Subject category	Particle Physics - Experiment
Accelerator/Facility, Experiment	CERN LHC ; ATLAS
Free keywords	opportunistic computing ; fine-grained data processing
Abstract	Continued growth in public cloud and HPC resources is on track to overcome the dedicated resources available for ATLAS on the WLCG. Example of such platforms are Amazon AWS EC2 Spot Instances, Edison Cray XC30 supercomputer, backfill at the Tier-2 and Tier-3 sites, opportunistic resources at the Open Science Grid, and ATLAS High Level Trigger farm between the data taking periods. Because of opportunistic resources specifics such as preemptive job scheduling and data I/O, their efficient usage requires workflow innovations provided by the ATLAS Event Service. Thanks to the finer granularity of the Event Service data processing workflow, the opportunistic resources are used more efficiently. We report on our progress in scaling opportunistic resource usage to double-digit levels in ATLAS production.