| dc.description.abstract | In recent years, with the development of technologies such as Software-Defined Networking, Network Function Virtualization and Infrastructure as Code (IaC), modern Information and Communication Technology infrastructures have become more software-based, providing faster deployment, scaling and simplified network management. Moreover, the number of organizations working with IaC is growing, typically consisting of highly skilled professionals with different educational and cultural backgrounds. In such organizations, effective collaboration between various human actors is crucial, and it often relies on efficient knowledge management.
Various human participants, including application developers, policy officers and network engineers, possess varying levels of understanding regarding the system they collaborate around. To develop a common language between them, we propose a knowledge-based approach, allowing formal representation of different concepts and their relationships. By adding semantics to the pre-deployment task of defining an infrastructure and representing concepts based on Description Logics, we develop a prototype interpretable by both humans and software agents.
In this thesis, we utilize the Semantic Web Technologies and follow the ontology development process to define an ontology to represent knowledge about Docker services and networks. We demonstrate the possible uses of our knowledge-based approach by creating knowledge graphs with data from several Docker Compose-based scenarios. Moreover, we apply the reasoning mechanism to check the consistency of each knowledge base and verify whether the definitions of web applications and two 5G Core Network solutions comply with a set of pre-defined logic rules. With our analysis of realistic IaC deployments, we confirm the potential of drawing new conclusions based on formal knowledge management and well-defined rules, which may assist various actors in making more informed decisions. | |
