Design and Deployment of a Test Infrastructure for Service Migration and Telemetry Enhancement
Terentev, Nikita (2025)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025060520805
https://urn.fi/URN:NBN:fi:amk-2025060520805
Tiivistelmä
This thesis investigates the implementation of a cloud-native, open-source infrastructure stack deployed on a bare-metal Kubernetes cluster. The study addresses the need for scalable, resilient, and observable platforms capable of supporting complex data and application workflows without relying on proprietary or cloud-hosted services.
The theoretical framework is based on principles of distributed systems, declarative infrastructure management, and automation. Emphasis is placed on Infrastructure as Code, templating, and declarative configuration to ensure reproducibility and maintainability of system components. The research draws from best practices in DevOps and Site Reliability Engineering (SRE), with a focus on fault tolerance, eventual consistency, and modular service design.
The results show that a fully open-source stack can be integrated into a coherent platform supporting storage, authentication, monitor-ing, logging, and metadata governance. While not aiming to compare directly with managed cloud services, the implementation demonstrates that open infrastructure is viable for complex workflows in controlled environments. Its successful orchestration highlights the practical and cost-efficient benefits of using open-source tools to build customizable, maintainable systems on bare-metal hardware.
The theoretical framework is based on principles of distributed systems, declarative infrastructure management, and automation. Emphasis is placed on Infrastructure as Code, templating, and declarative configuration to ensure reproducibility and maintainability of system components. The research draws from best practices in DevOps and Site Reliability Engineering (SRE), with a focus on fault tolerance, eventual consistency, and modular service design.
The results show that a fully open-source stack can be integrated into a coherent platform supporting storage, authentication, monitor-ing, logging, and metadata governance. While not aiming to compare directly with managed cloud services, the implementation demonstrates that open infrastructure is viable for complex workflows in controlled environments. Its successful orchestration highlights the practical and cost-efficient benefits of using open-source tools to build customizable, maintainable systems on bare-metal hardware.