Assurance de la qualité de service avant le déploiement des systèmes fog avec l’ingénierie basée sur des modèles et DevOps

Fog Computing decentralizes the Cloud by bringing computation, storage, and network services closer to the edge. This reduces latency and bandwidth usage while improving real-time processing. However, the complexity and heterogeneity of Fog systems, often comprising diverse entities, make lifecycle management challenging and costly. Runtime error handling frequently requires revisiting earlier phases, which is both timeconsuming and expensive. Ensuring reliability through pre-deployment verification is therefore essential. Fog systems are deployed in domains such as healthcare, automotive, and smart cities, further complicating verification and deployment processes. To address these challenges, we propose a generic and customizable approach based on a two-step verification process. This approach focuses on the design-time and pre-deployment phases, automating key verification and deployment activities. Our solution features a customizable Fog modeling language, design-time verification of non-functional properties (e.g., security, energy), preparation of pre-deployment configurations, and integration with industrial DevOps tools and Quality of Service (QoS) solutions. By combining Model-Based Engineering and DevOps practices, our approach ensures QoS, reduces deployment costs, and enhances automation to tackle the complexity of Fog systems. We validated this approach using three literature-based use cases—smart campus, smart parking, and smart hospital. Results demonstrate its effectiveness in QoS verification, deployment automation, and reducing complexity and costs, highlighting its relevance to state-of-the-art engineering and DevOps practices.