Orquestração de contêineres em computação de borda: uma avaliação de desempenho comparativa entre K3s e K0s

Abstract

The rising adoption of containers as a lightweight virtualization technology has paralleled the popularity of microservice architectures, given containers’ inherent scalability, isolation, and ephemeral nature. While container orchestration frameworks have traditionally been utilized in data centers, there is growing interest in deploying these frameworks on resource-constrained devices, such as Internet of Things (IoT) devices, edge gateways, and developer workstations. This shift is driven by the need for automated provisioning, deployment, scaling, networking, and load balancing across diverse environments. However, with a range of container orchestration solutions available, selecting the appropriate platform remains a challenge for developers. This paper presents a comparative analysis of container orchestration platforms, beginning with a literature reviewto examine various platforms and their applicability across different environments. Following this, K3s and KOs are selected for an in-depth evaluation under a specified scenario, focusing on their minimal resource requirements, as well as control plane. Through this approach, we identify key factors and optimal use cases, offering insights into the suitability of these platforms for resource-constrained hardware environments.