Shared Data Layer on Kubernetes Cloud Platform for Subscriber Data Management

Abstract

In the era of cloud-native technologies, the deployment and management of network functions have evolved from traditional Virtual Network Functions (VNF) to more efficient and scalable Cloud-Native Functions (CNF). The analysis emphasizes the significance of container orchestration in improving resource efficiency, scaling, and deployment agility, especially in environments demanding high availability and real-time data access. This paper investigates the deployment of a Shared Data Layer (SDL) with CNF over VNF for Subscriber Data Management (SDM) in a Kubernetes environment. The research aims to design and implement a cloud-native SDM solution with GUI based Docker image and Helm charts for manual deployments. This work includes building and verifying Docker images, deploying Helm charts, and assessing deployment status through pod monitoring. Performance evaluation shows CNFs significantly outperform VNFs in resource utilization, scalability, and deployment speed. CNF deployments completed in approximately 1 hour 30 minutes, a 47% reduction compared to VNF deployments, which took around 2 hours 49 minutes. The SDL's integration further ensures robust performance and data consistency. Results confirm that CNFs, combined with SDL, offer superior operational agility and scalability rather than VNFs for modern network infrastructures.