• declarative programming
• cloud-iot continuum
• fog computing
• logic programming
• orchestrators
• continuous reasoning
• cloud-iot infrastructures
• microservices
• multi-service applications

Designing platforms and methodologies for the orchestration and management of multi-service applications in the Cloud-IoT continuum is becoming more and more a problem of primary importance, due to the current paradigm shifts in the applications development and the new computing paradigms. In light of these needs, in this thesis, we study and implement innovative orchestration techniques, based on the Continuous Reasoning approach, to achieve a scalable, continuous and QoS-compliant management of multi-service applications on large-scale, geographically distributed Cloud-IoT networks. We propose a declarative methodology, and its prototype FogBrainX, to speed up QoS-aware, decision-making in Cloud-IoT application management, and we design and develop a next-generation orchestrator prototype, FogArm, to perform continuous and QoS-compliant management of multi-service applications in Cloud-IoT settings, in continuity with either CI/CD pipelines and infrastructure distributed monitoring tools. Finally, both the methodology and the orchestrator are extensively assessed. To the best of our knowledge, FogArm represents a first complete prototype of a next-gen orchestrator for the continuous QoS-compliant management of multi-service applications on geographically distributed Cloud-IoT nodes.