Cloud environment assessment using clustering techniques on microservices dataset

Abstract

The telecommunication sector is among the fast-growing industries with technological breakthroughs each couple of years. The main goals for businesses are customer experience improvement and lowering the overall costs of service delivery. In these settings, the business operations team confront increasingly more challenges in various forms: fast delivery protocols for new services, better support for production services and increase the value of the enterprise. Deployment environment assessment is paramount in the digital transformation of organizations all over the industry, not only because it ensures service continuity, but also because this could control the cloud resources costs. This paper proposes a solution that can be integrated in a smart automation architecture: unsupervised clustering for microservices deployment environment evaluation. The present paper presents a telecommunication setup, although the results are field-independent. The results section of the paper presents an evaluation of various unsupervised clustering models on public microservice time-series dataset.