Segmented Network Architecture for Promoting High Availability in Fog Computing through Middleware

Abstract

This paper proposes an architecture for deploying applications on a fog computing environment by adding another layer of fog nodes in a network segment that gains high software application availability. The conventional fog computing architecture would permanently shift the storage, applications, and data from cloud servers to fog nodes, thus reducing the dependency on the cloud. As a result, fog nodes are burdened with the task previously done by cloud servers and have become "mini cloud servers." Instead of permanently shifting the tasks from cloud servers to fog nodes, the proposed architecture would only do the shifting, when necessary, like if an internet outage. Additionally, this research also introduced the middleware application that acts as a detector and replacement if service outage so that the availability of the services is not interrupted, especially during the internet outage, by adding another layer of fog node in a network segment. The computational process occurs between end-users and the fog nodes without having to rely on cloud servers. An experiment was conducted to test the performance of the proposed architecture. From the experiment, it can be concluded that the deployment of fog nodes in a segmented network is possible and able to increase the availability of data and services if an internet outage.

Real-Time Bus Arrival Time Estimation API using WebSocket in Microservices Architecture

“…All communication between machines is done through the API, but currently, the word "API" refers more to a web-based API [15,16,17]. Web APIs, often called Web Service [18], are usually used as the back-end to build applications with various platforms, such as web-based and mobile applications because the API is designed not to be tied to the platform type of client that could access it [19,20]. One of the most widely used API types is REST because of its simple scheme [21,22,23].…”