Feasibility of Fog Computing Deployment based on Docker Containerization over RaspberryPi

Abstract

Fog computing is strongly emerging as a relevant and interest-attracting paradigm+technology for both the academic and industrial communities. However, architecture and methodological approaches are still prevalent in the literature, while few research activities have specifically targeted so far the issues of practical feasibility, cost-effectiveness, and efficiency of fog solutions over easily-deployable environments. In this perspective, this paper originally presents i) our fog-oriented framework for Internet-of-Things applications based on innovative scalability extensions of the open-source Kura gateway and ii) its Docker-based containerization over challenging and resource-limited fog nodes, i.e., RaspberryPi devices. Our practical experience and experimental work show the feasibility of using even extremely constrained nodes as fog gateways; the reported results demonstrate that good scalability and limited overhead can be coupled, via proper configuration tuning and implementation optimizations, with the significant advantages of containerization in terms of flexibility and easy deployment, also when working on top of existing, off-the-shelf, and limited-cost gateway nodes.