N. V. Greben1, D.А. Oleynik2, A.K. Kiryanov3, A. A. Sychugov4, E.I. Minakov5

1,2,3 Joint Institute for Nuclear Research (Dubna, Russia)

3 NRC «Kurchatov Institute» − PNPI (Dubna, Russia)

3 Tula State University (Dubna, Russia)

4,5 Tula State University (Tula, Russia)

1 ngreben@jinr.ru, 2 danila.oleynik@jinr.ru, 3 kiryanov_ak@pnpi.nrcki.ru, 4 xru2003@list.ru, 5 eminakov@bk.ru

Problem Statement. Modern research in natural sciences is characterized by a significant volume of produced data. The high-energy physics experiments, where data volumes can reach tens of petabytes, leads on in this field. Specialized computing systems designed for processing such vast amounts of data are typically problem-specific and their development and configuration are complex tasks requiring significant resources and time. One of example is the SPD experiment at the NICA collider, which imposes special requirements on the comprehensive primary processing of collected data. The development of a specialized computing system necessitates the execution of various debugging tests utilizing dozens or even hundreds of compute nodes. Manual configuration of required computing infrastructure is highly time-consuming, prone to errors, and non-deterministic. Consequently, the automation of this process is imperative.

Goal. Reducing costs and manual operations when deploying computing clusters in cloud infrastructure.

Results. This paper presents a methodology and implementation for the automated deployment of computing infrastructure within a private oVirt cloud, utilizing the Ansible configuration management platform. To facilitate rapid deployment, removal, and updating of compute node pools, a command-line interface (CLI) application with state storage in a database has been developed.

Practical significance. This approach has been validated through testing on a prototype of the hardware component for the SPD experiment's primary data processing system – the SPD Online Filter. The methodology is adaptable and can be generalized for application in similar deployment procedures for computing systems in other projects.

The work was carried out with the financial support of a Grant in the form of a research grant from the Committee on Science and Innovation of the Tula Region No. 15 dated June 21, 2024.

Greben N. V., Oleynik D.А., Kiryanov A.K., Sychugov A. A., Minakov E.I. Automation of computing cluster deployment in cloud infrastructure // Radioengineering. 2026. V. 90. № 9. P. 142−149. DOI: https://doi.org/10.18127/j00338486-202609-15

1. Abazov V. et al. Technical design report of the spin physics detector at NICA. Natural Sci. Rev. 2024. V. 1. № 1; https://doi.org/10.54546/NaturalSciRev.100101.
2. Andreev V., Belova A., Galoyan A., Gerassimov S., Golovanov G., et al. Offline Software and Computing for the SPD Experiment e-Print: 2111.02047 [physics.ins-det]; https://arxiv.org/abs/2111.02047.
3. Sadiku M.N.O., Ezeand K.G., Musa S.M. High-throughput computing. Int. J. Trend Res. Dev. 2018. № 54. Р. 329–330.
4. oVirt project. https://www.ovirt.org.
5. Dakic Vedran, Chirammal Humble Devassy, Mukhedkar Prasad. Mastering KVM virtualization. Second Edition. Birmingham. UK: Packt Publishing. 2020. 686 р.
6. Amundsen M. RESTful Web API patterny i praktiki. Kazahstan: Sprint Book. 2025. 464 s. (in Russian).
7. Open Source Client for Enterprise Identity Management https://sssd.io/.
8. Ansible project. Red Hat; https://www.ansible.com.
9. Morris K. Programmirovanie infrastruktury. Izd. 2-e. SPb: BHB. 2024. 416 s. (in Russian).
10. Rahman A., Mahdavi-Hezaveh R., Williams L. A systematic mapping study of infrastructure as code research // Information and Software Technology. 2019. V. 108. Р. 65−77; https://doi.org/10.1016/j.infsof.2018.12.004.
11. Cloud-Init documentation. https://cloudinit.readthedocs.io.
12. Owens M., Grant A. The definitive guide to Sqlite. London. UK: Springer. 2010. 350 р.
13. The Rust Programming Language. https://doc.rust-lang.org/book/.