Declarative application and network management in the Cloud-Edge continuum

The proliferation of Internet of Things (IoT) devices and latency-sensitive applications exposes the limits of Cloud-centric models. This thesis investigates declarative methodologies for managing applications and networks across the Cloud-Edge continuum, focusing on data-aware placement, deployment evaluation, and intent-based orchestration. We propose a declarative mathematical model for data-aware application placement that integrates data properties into placement and routing decisions. Implemented in DAPlacer and EdgeWise, it leverages Prolog-based reasoning for dynamic multi-service allocation while satisfying heterogeneous requirements. To enhance adaptivity, continuous reasoning mechanisms reduce computation time during migrations. We also introduce ECLYPSE, a simulation and emulation framework for evaluating Cloud-Edge placement strategies under realistic, reproducible conditions, enabling analysis of resource allocation, performance, and orchestration policies. Furthermore, we advance Intent-Based Networking (IBN) through declarative methods for modelling, translation, and conflict resolution of high-level intents. The DIPS and MultiDIPS prototypes enable scalable orchestration of Virtual Network Function (VNF) chains, while dgLBF provides declarative traffic engineering by translating network-level intents into reliable, latency-aware forwarding configurations.