Публікація:
Adaptive methods for managing distributed computing with built-in selfhealing mechanisms

dc.contributor.authorVolk, M. O.
dc.contributor.authorBuhrii, A. M.
dc.contributor.authorZaihraev, D. S.
dc.contributor.authorPanchenko, Y. O.
dc.contributor.authorZveriev, P. V.
dc.date.accessioned2025-10-16T16:53:48Z
dc.date.available2025-10-16T16:53:48Z
dc.date.issued2025
dc.description.abstractCloud computing systems increasingly face challenges of heterogeneity, scalability, and reliability. This paper presents an integrated approach to distributed computing management with embedded self-recovery mechanisms in heterogeneous cloud environments. The architecture combines adaptive scheduling, predictive analytics, and machine learning methods (LSTM, Random Forest, reinforcement learning) to forecast failures and optimize recovery scenarios. Experimental evaluation demonstrates improved fault prediction accuracy (12–15%), reduction of false alarms (25–32%), faster incident response (from 11–17 minutes to 5–9), and a 17% decrease in operational costs. The results confirm the technical and economic feasibility of the proposed approach for missioncritical cloud infrastructures.
dc.identifier.citationAdaptive Methods for Managing Distributed Computing with Built-in Self-Healing Mechanisms / M. Volk et al. // Computer and information systems and technologies : Proceedings of Eighth International Scientific and Technical Conference, 9-10 October 2025. – Kharkiv : NURE, 2025. - P. 5.
dc.identifier.urihttps://openarchive.nure.ua/handle/document/32971
dc.language.isoen
dc.publisherNURE
dc.rightsAttribution-NoDerivatives 4.0 Internationalen
dc.rights.urihttp://creativecommons.org/licenses/by-nd/4.0/
dc.subjectcloud computing
dc.subjectdistributed systems
dc.subjectpredictive analytics
dc.subjectselfrecovery
dc.subjectmachine learning
dc.subjectresource management
dc.subjectheterogeneous environments
dc.subjectservice availability
dc.titleAdaptive methods for managing distributed computing with built-in selfhealing mechanisms
dc.typeConference proceedings
dspace.entity.typePublication

Файли

Оригінальний пакет
Зараз показано 1 - 1 з 1
Завантаження...
Зображення мініатюри
Назва:
5-6_unlocked.pdf
Розмір:
693.58 KB
Формат:
Adobe Portable Document Format
Ліцензійний пакет
Зараз показано 1 - 1 з 1
Немає доступних мініатюр
Назва:
license.txt
Розмір:
10.74 KB
Формат:
Item-specific license agreed upon to submission
Опис: