The German university based WLCG Tier-2 centres successfully contributed a significant fraction of computing power for the Runs 1-3 of the LHC. But for the upcoming Run 4, with its increased need for computing resources for the various HEP computing tasks in terms of storage and computing power, there is a necessity for a transition to a new model. In this context, the German community under the FIDIUM project is making interdisciplinary resources of the National High Performance Computing (NHR) usable within the WLCG and centralising mass storage at the Helmholtz centres. At the Goettingen campus, there is both a WLCG Tier-2 site (GoeGrid) and a large HPC cluster Emmy by NHR and the North German Supercomputing Alliance (HLRN). The integration is done by virtually extending the GoeGrid batch system with containers, turning the HPC nodes into virtual worker nodes with their own partitionable job scheduling in order to run the HEP jobs which serve the ATLAS collaboration in the case of GoeGrid. Submission and management of these containers is automated using COBalD (the Opportunistic Balancing Daemon) and TARDIS (The Transparent Adaptive Resource Dynamic Integration System). Data is provided via the GoeGrid mass storage for which a dedicated network connection has been established. Continuous production of the ATLAS jobs is currently tested in a one-year pilot phase. The setup, experiences, performance tests and an outlook are presented.
Łakomiec et al. (Tue,) studied this question.