The management of modern networks, especially Software-Defined Networks (SDN), is becoming increasingly complex due to the scale, heterogeneity, and dynamic nature of network topologies and traffic patterns. Traditional machine learning and optimization techniques often struggle to capture the intricate relationships and dependencies inherent in network structures. Graph Neural Networks (GNNs), a cutting-edge deep learning paradigm designed to operate directly on graph-structured data, offer unprecedented capabilities to model and analyze network topologies, device interactions, and traffic flows.