| Abstract: |
The rapid shift from centralized cloud computing to the decentralized edge-cloud continuum has resulted in a pressing need for a realistic environment in which distributed systems can be tested and analyzed. Many existing laboratory environments are either hardware-dependent, difficult to scale, or lack the ability to integrate necessary orchestration and monitoring functions to support the evaluation of edge cloud strategies. This paper addresses the research question of how to design a scalable and reproducible interactive laboratory that enables controlled experimentation across the edge-cloud continuum. We demonstrate the design and implementation of an interactive laboratory environment that incorporates physical edge nodes, a virtualized private cloud infrastructure, container-based orchestration, and infrastructure as code, showing that this combination provides improved configurability and reproducibility over traditional testbeds. The platform also includes a web-based interactive dashboard that facilitates real-time deployment control, monitoring, and visualization of performance metrics, including including latency, bandwidth usage, energy consumption, and fault tolerance analysis. Experimental case studies for validating the laboratory through IoT data streaming services and distributed artificial intelligence inference with dynamic edge offloading strategies were also performed. The validation environment also includes IoT workload simulators for emulating realistic IoT data streams for controlled experimentation under various system conditions. The results verify the effectiveness of multi-node deployment, full experimental reproducibility, and improvements in latency for edge-enabled scenarios compared with cloud-only scenarios. Moreover, the platform reveals efficiency trade-offs consistent with prior research and its suitability as a research and education platform for edge-integrated distributed systems. |