Performance Analysis of Data Storage and Management Models for Microservice-Based Edge Systems

Abstract

The rapid proliferation of distributed cloud applications has increased demand on computing infrastructures. Thus, it is important to reduce unnecessary data flow to cloud services to decrease costs and system latency for end-users. Accordingly, edge computing can provide significant benefits in optimizing data traffic. In terms of performance, heavyweight applications are undesired in edge systems. Instead, microservices are preferred for providing lightweight and suitable solutions for edge computing systems with limited resources. Importantly, the data storage and management mechanism used by a microservice-based edge system is a vital consideration for optimizing data traffic between the cloud and the end-user system. Ensuring data reliability and edge system performance depends on the adopted model. This study examined three different data storage and management models for microservice-based edge systems. The models included microservices with their own databases or where a database was distributed among microservices. All models were tested for synchronous and asynchronous operation of microservices in distributed edge devices. The study aim was to find a suitable data storage and management model, especially for microservice-based edge systems with limited resources. © 2022 IEEE.