Karlsruhe 2024 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 17: Data, AI, Computing 2 (analysis tools)

T 17.7: Vortrag

Montag, 4. März 2024, 17:30–17:45, Geb. 30.34: LTI

AUDITOR: Accounting for opportunistic resources — •Raghuvar Vijayakumar, Michael Böhler, Benjamin Rottler, Dirk Sammel, and Markus Schumacher — Universität Freiburg

In response to the increasing demand for computing resources in High Energy Physics (HEP), we have integrated under-utilized computing resources using COBalD/TARDIS in an opportunistic manner. However, resource sharing requires robust accounting. We present AUDITOR (AccoUnting DatahandlIng Toolbox for Opportunistic Resources), a flexible and extensible ecosystem tailored to address a broad range of use cases and infrastructures. AUDITOR employs specialised collectors that monitor, capture and record accounting data, subsequently stored in a database. The recorded data are then made accessible to plugins, which analyse the accounting information to perform specific tasks, such as computing the CO2 footprint and forwarding the accounting data to other accounting systems. AUDITOR is written in Rust, a modern programming language optimised for performance, type safety and concurrency. The python client pyauditor allows to utilize all AUDITOR functionalities via a python interface. In this talk, we present how AUDITOR works and how Pyauditor can be used to query and analyze large accounting data sets.

Keywords: AUDITOR; Computing Resources; High Energy Physics (HEP); COBalD/TARDIS