USING DISTRIBUTED CLOUDS for SCIENTIFIC COMPUTING

ABSTRACT

Nowadays, cloud resources are the most flexible tool to provide access to infrastructures for establishing services and applications. However, it is also a valuable resource for scientific computing. At the Joint Institute for Nuclear Research, the computing cloud was integrated with the DIRAC system. It allowed for the submission of scientific computing jobs directly to the cloud. Thanks to the experience, the cloud resources of several organizations from the JINR Member States were integrated in the same way. It increased the total amount of cloud resources accessible in a uniform way through DIRAC, in the scope of the so-called Distributed Information and Computing Environment (DICE). Folding@Home tasks related to the SARS-CoV-2 virus were submitted to all available cloud resources. In addition to useful scientific results, such experience was also helpful in obtaining information about the performance, limitations, strengths, and weaknesses of the combined system. Based on the gained experience, the DICE infrastructure was tuned to successfully perform real user jobs related to Monte-Carlo simulation for the Baikal-GVD experiment. Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).