EUVIMEDEUVIMED entdecken
Crossref

Proactive Auto-Scaling of Kubernetes Services Based on Machine Learning

A.S. Pazinin

Elektronne modelûvannâ · 27.04.2026 · 48 (2) · 69-86

DOI: 10.15407/elmodel.48.02.069 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Elektronne modelûvannâ, online erschienen am 05.05.2026. Verfasst von A.S. Pazinin. Die Arbeit nennt 20 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To optimize the use of computing resources, reactive and proactive approaches to service au-toscaling in Kubernetes are considered: the standard reactive Horizontal Pod Autoscaler (HPA) and a proactive autoscaler based on machine learning using LSTM. A controller has been de-veloped and proposed that collects CPU metrics from Prometheus, trains and updates the mod-el, predicts short-term load dynamics, and adjusts the number of replicas via the Kubernetes API. Metrics for predictions and decisions are sent to Pushgateway and visualized in Grafana. Experimental studies in an Azure Kubernetes Service cluster with controlled container load showed a 30 % reduction in total vCPU usage compared to HPA while maintaining the same service level, reducing scaling latency (scaling up in 30-60 s versus 75-90 s; shrink time of 60-90 s versus 90-150 s) and elimination of “jitter.” The results confirm the effectiveness of ap-plying proactive Kubernetes service autoscaling based on machine learning methods for ser-vices with stable or seasonal traffic patterns.

Zitieren

APA

Pazinin, A. (2026). Proactive Auto-Scaling of Kubernetes Services Based on Machine Learning. Elektronne modelûvannâ, 48(2), 69-86. https://doi.org/10.15407/elmodel.48.02.069

Vancouver

Pazinin A. Proactive Auto-Scaling of Kubernetes Services Based on Machine Learning. Elektronne modelûvannâ. 2026;48(2):69-86. doi:10.15407/elmodel.48.02.069

BibTeX
@article{Pazinin2026,
  author = {Pazinin, A.S.},
  title = {Proactive Auto-Scaling of Kubernetes Services Based on Machine Learning},
  journal = {Elektronne modelûvannâ},
  year = {2026},
  volume = {48},
  number = {2},
  pages = {69--86},
  doi = {10.15407/elmodel.48.02.069}
}
Kostenlos starten →