The widespread adoption of the cloud continuum paradigm poses significant challenges such as the management of heterogeneous infrastructural devices, the strict security and privacy requirements, and the complex data governance constraints. While the cloud provides access to advanced services that are often inaccessible to small and medium organizations, edge and fog resources are essential to meet latency, locality, and efficiency demands. The main benefits provided by the cloud continuum is to provide scalability, flexibility, and resilience by seamlessly integrating networking, storage, and computing resources across different layers. In this context, we present a lightweight agent, coined EdgeGuard, designed for seamless integration into heterogeneous infrastructures within a computing continuum architecture. It enables real-time monitoring of multiple metrics (e.g., resource utilization, energy consumption) and provides predictive capabilities to anticipate and mitigate potential issues before they escalate. We validate our proposal through an experimental scenario involving a diverse set of infrastructural devices distributed across the continuum with experts in the field. The evaluation shows that EdgeGuard consistently outperforms human experts across all measured metrics. These results highlight its effectiveness in proactive monitoring and correction of infrastructural issues, making it a suitable tool for modern distributed computing environments. Ultimately, EdgeGuard contributes to building more resilient, scalable, and intelligent systems within the evolving landscape of edge-cloud continuum.
Bonilla et al. (Thu,) studied this question.