Seamless Task Offloading on Multi-Clouds and Edge Resources: an Experiment

Cloud computing has been growing at an increasing rate over the last few years. Commercial and scientific

applications have come to rely on it as a development tool due to its exceptional characteristics in processing power and storage.

The trend has been augmented with the emergence of the Internet of Things and smart processing devices at the edge.

Contrary to the line of thought commonly adopted, we present in this work an alternative platform that considers edge devices as

possible processing nodes, and provide a two-level task scheduling deployment that can handle not only binary modules,

but also code-level fragmentations. We also go through a simple implementation of the platform, using production-ready

solutions, while validating it on public and private clouds, and physically-separated edge devices.