Applying Patterns to Support Deployment in Cloud-Edge Environments: A Case Study

Abstract

A major trend followed by IT experts and Software developers in recent years is represented by the “Cloudification” of existing applications, with a strong shift of computations and data from local and centralized servers to remote, distributed data-centers. Indeed, using Cloud resources has reduced, for most SMEs, both the initial investments in hardware and software assets and maintenance costs, making it a viable choice in many situations. On the other hand, Cloud Computing requires to store consistent volumes of data on remote databases, with a series of consequences on data privacy that need to be carefully addressed. Moreover, the advent of the Internet of Things, with the huge quantity of data that smart devices continuously produce and consume, often in real time, renders the transfer of information to and from remote servers too cumbersome, as it relies on network speed and continuous availability. New programming paradigms have thus emerged, such as Cloud-Edge, which tries to combine benefits deriving from the exploitation of the resources offered by Cloud architecture and the need to consume data locally. The Cloud-Edge paradigm requires a careful design of the integration between Cloud and Edge architectures, in order to avoid bottlenecks and efficiently exploit both local and remote resources. In this paper a methodology based on Architectural, Computational and Deployment Patterns will be presented to support the deployment of applications in Cloud-Edge environments, starting from pre-existing software solutions.