From Edge to Cloud: A Helm-Based Installer for Deploying a Multi-Tenant Application and Services on Kubernetes

  • Tobias Fischer

    Student thesis: Master's Thesis

    Abstract

    The digitalization of the energy sector and the accessibility of near real-time energy data are essential for enabling the energy transition. The European Union’s EDDIE project addresses this need by providing a central interface for energy data access and a marketplace for data-driven services. At the core of this infrastructure is AIIDA, a software component that collects and processes near real-time data from devices such as smart meters, solar inverters, and batteries, and provides a secure interface for service integration. This thesis presents a Helm-based installer for deploying AIIDA and its services in both cloud and edge environments using Kubernetes. The installer manages the complete lifecycle of AIIDA and its associated services, including installation, upgrades, configuration, and removal. It is implemented as a Go-based application that interacts with Kubernetes, Helm, and SealedSecrets, and is accessible only through the AIIDA interface via REST and UI components. Special emphasis is placed on secure multi-tenant deployments, service isolation, and edge compatibility. The installer includes a generic setup script for simplified deployment on any Linux system and supports Helm chart-based service distribution through a centralized repository. Limitations of existing GitOps-based deployment solutions are addressed by avoiding reliance on cloud-based infrastructure. Challenges encountered during development include managing identity and access using Keycloak in dynamic environments and securing MQTT communication. The proposed solution demonstrates that a decentralized, service-oriented energy data platform can be effectively deployed across heterogeneous infrastructure environments, laying the groundwork for further marketplace integration and automated service management.
    Date of Award2025
    Original languageEnglish
    SupervisorChristoph Schaffer (Supervisor)

    Studyprogram

    • Mobile Computing

    Cite this

    '