Cluster for artificial intelligence research in an educational institution with differentiated access rights
Objective. The goal of the study is to present a cluster configuration that effectively manages infrastructure resources and security for artificial intelligence training in educational organizations. Method. The study is based on methods of systems analysis, information processing, artificial intelligence. Result. An architectural solution is proposed, based on the synergy between a lightweight Kubernetes distribution (K3s) and the multi-user platform JupyterHub. The study examines the architectural stack, access control mechanisms, traffic routing, as well as strategies for monitoring and ensuring fault tolerance. Conclusion. The configuration strikes a balance between competing requirements: control plane fault tolerance (ensuring a quorum even if one node fails), resource efficiency (minimizing overhead by replacing etcd with SQLite), and operational simplicity (a single point of configuration via Helm charts).
Authors
- О. В. Ланкин (ORCID: https://orcid.org/0009-0001-9751-0458)
- Anton Dmitrievich Popov (ORCID: https://orcid.org/0000-0002-6583-102X)
- A. E. Dmitriev
Institutions
- Voronezh Institute of Ministry of Internal Affairs (RU)
Publication Details
- Journal
- Herald of Dagestan State Technical University Technical Sciences
- Published
- 2026-11-12
- DOI
- https://doi.org/10.21822/2073-6185-2026-53-3-112-118
- Primary Topic
- Cloud Computing and Resource Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00