Automation and provisioning of Kubernetes on bare-metal telco edge infrastructures

Cloud computing has revolutionized the deployment of innovative services, with cloud-native applications becoming a fundamental design pattern. Kubernetes has emerged as the leading platform for container and application management in this paradigm. For telecommunication companies (Telcos) with numerous edge sites, adopting cloud-native solutions offers the potential for rapid deployment, configuration, and maintenance of network applications, while ensuring a robust certification process. However, this shift requires a certified, modular, and extensible framework to manage multiple edge sites along with their associated Network Functions and services. The Sylva Project aims to provide such a framework, offering a reference implementation based on open-source software tailored to European Telcos' needs. Sylva seeks to automate and standardize the traditionally manual and fragmented process of deploying and managing edge sites with bare-metal or virtualized nodes. The automation of bare-metal Sylva deployments presents a unique challenge, requiring deep, domain-specific knowledge of the underlying infrastructure. This thesis proposes an architecture based on the Server Operator and its related Custom Resource Definitions (CRDs). An important component of this solution is to seamlessly integrate within the existing stack, while following the patterns and standards of Kubernetes and the tools used within Sylva. This is achieved by extending Sylva's current operators to simplify the configuration of bare-metal nodes. By doing so, the proposed architecture aims to enhance collaboration between infrastructure administrators and application developers, ultimately increasing deployment agility in Telco environments.