Automating resource monitoring and visualization for microservices in Kubernetes using VPA and custom controllers

Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

School of Electrical Engineering | Master's thesis

Authors

Date

2025-07-30

Department

Major/Subject

Cloud and Network Infrastructures

Mcode

Degree programme

Master's Programme in ICT Innovation

Language

en

Pages

64

Series

Abstract

As organizations adopt microservices and container-based architectures, Kubernetes has become the leading platform for orchestrating workloads. Since the complexities introduced by microservices has emerging importance in industry, efficiently managing resources in Kubernetes, especially in on-premises environments is critical. This thesis investigates current resource management practices, based on the research results, designed and implemented a practical tool for real-time monitoring and visualization. The solution integrates with exisiting tools to provide automated resource tracking, estimation, and visualization, supporting data-driven decision-making while requiring minimal operational overhead for adoption. The tool was evaluated through real-world deployment and feedback from practitioners, demonstrating its effectiveness in improving resource observability, helping avoid over- and under-provisioning, and eventually reducing operational costs. Although limitations presents, the results show real benefits for the industry.

Description

Supervisor

Manner, Jukka

Keywords

Kubernetes, microservices, autoscaling, resource management, controller, monitoring

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202508196569

Collections

[dipl] Sähkötekniikan korkeakoulu / ELEC