Artifact for Service-Level Energy Modeling and Experimentation for Cloud-Native Microservices

TL;DR

This work addresses the lack of service-level energy measurement for cloud-native microservices by introducing GOXN, a Kubernetes-native energy experimentation engine that integrates per-container energy data (via Kepler) and network/storage energy proxies to yield per-service energy. It employs a component-level additive model, $E_{service} = E_{compute} + E_{network} + E_{storage}$, where $E_{compute}$ comes from Kepler and $E_{network}$/$E_{storage}$ come from proxy metrics with energy-intensity factors. The artifact includes an executable framework and a replication package, offering end-to-end (Path A) and data-only (Path B) reproduction routes, along with guidance for extension and provenance. Key findings show that excluding network and storage can underestimate auxiliary-service energy by up to 63%, and that heavy tracing shifts energy dominance toward network and storage, highlighting the importance of service-level energy accounting in cloud-native deployments.

Abstract

Recent advancements enable fine-grained energy measurements in cloud-native environments (e.g., at container or process level) beyond traditional coarse-grained scopes. However, service-level energy measurement for microservice-based applications remains underexplored. Such measurements must include compute, network, and storage energy to avoid underestimating consumption in distributed setups. We present GOXN (Green Observability eXperiment eNginE), an energy experimentation engine for Kubernetes-based microservices that quantifies compute, network, and storage energy at the service level. Using GOXN, we evaluated the OpenTelemetry Demo under varying configurations (monitoring, tracing, service mesh) and steady synthetic load, collecting metrics from Kepler and cAdvisor. Our additive energy model derives service-level energy from container-level data. Results show that excluding network and storage can underestimate auxiliary-service energy by up to 63%, and that high tracing loads shift energy dominance toward network and storage.