HEPSCORE Integration for WLCG Accounting

Abstract

This thesis investigates the enhancement of the CPU accounting workflow in the Worldwide LHC Computing Grid (WLCG) at CERN. The study focuses on analyzing the existing environment, understanding benchmarking metrics like HEPSCORE, and proposing an integrated approach for incorporating new benchmarks. The research highlights the need for updating Spark jobs code to include HEPSCORE values and streamline reporting to the third-party service, APEL. By collaborating with colleagues and utilizing a dedicated testing environment, successful integration is achieved. The findings emphasize the importance of documentation and knowledge sharing for sustained improvements in complex technical environments.