Automatically Scaling Applications with Helm and Kubernetes

Abstract

In today's dynamic cloud environments, modern applications must be able to scale automatically to meet fluctuating workloads without compromising performance or cost-efficiency. This thesis investigates the principles and practices of automatic application scaling using Helm and Kubernetes. The study explores theoretical scaling concepts and their real-world application. It uses a practical case study based on the Paperless-NGX document management system to demonstrate real-world application of scaling strategies. Key technologies such as Horizontal and Vertical Pod Autoscalers, the Cluster Autoscaler, Helm chart templating, and Kubernetes configuration management are explored in depth. Performance metrics, service dependencies, resource optimization, and persistent storage are analyzed to highlight challenges and solutions in scaling containerized applications. The findings suggest that a combined approach using Docker Compose and Kubernetes can significantly improve the efficiency, agility, and resilience of cloud-native systems.