Dynamic Distribution of High-Rate Data Processing from CERN to Remote HPC Data Centers

• DOI: 10.1007/s41781-020-00052-w
• Publication Date: 2021-02-09T22:52:05Z
• Authors: T. Boccali; D. Cameron; N. Cardo; D. Conciatore; A. Di Girolamo; G. Dissertori; P. Fernandez; A. Filipcic; M. Gila; C. Grab; J. Elmsheuser; V. Jankauskas; A. Klimentov; D. Kovalskyi; S. Lammel; D. Petrusic; T. C. Shulthess; F. G. Sciacca; C. Serfon; R. Walker; C. Wissing
• Keywords: Provisioning

Abstract The prompt reconstruction of the data recorded from Large Hadron Collider (LHC) detectors has always been addressed by dedicated resources at CERN Tier-0. Such workloads come in spikes due to nature operation accelerator and special high load occasions experiments have commissioned methods distribute (spill-over) a fraction sites outside CERN. present work demonstrates new way supporting Tier-0 environment provisioning elastically for such spilled-over workflows onto Piz Daint Supercomputer CSCS. This is implemented using containers, tuning existing batch scheduler reinforcing scratch file system, while still standard Grid middleware. ATLAS, CMS CSCS jointly run selected on up several thousand cores into shared environment, thereby probing viability performance computer site as demand extension Tier-0, which could play role addressing future LHC computing challenges luminosity LHC.