Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

An Open-Source Experimentation Framework for the Edge Cloud Continuum

Georgios Koukis, Sotiris Skaperas, Ioanna Angeliki Kapetanidou, Vassilis Tsaoussidis, Lefteris Mamatas

TL;DR

The paper presents CODECO, an open-source Experimentation Framework designed to accelerate rapid, cross-layer testing of Kubernetes-based edge cloud deployments. It introduces a declarative Experiment Descriptor and a microservice architecture that orchestrates cluster provisioning, application deployment, experiment execution, and results processing across diverse edge environments. The authors demonstrate three proof-of-concept scenarios: benchmarking network plugins across edge-oriented Kubernetes distributions, automated deployment of EdgeNet, and online anomaly-detection workflows tailored for edge systems. These contributions enable reproducible, end-to-end experimentation with configurable cross-layer parameters, supporting integration with external testbeds and edge platforms. The framework advances practical evaluation of edge orchestration, networking, and AI-driven workflows in a scalable, extensible manner.

Abstract

The CODECO Experimentation Framework is an open-source solution designed for the rapid experimentation of Kubernetes-based edge cloud deployments. It adopts a microservice-based architecture and introduces innovative abstractions for (i) the holistic deployment of Kubernetes clusters and associated applications, starting from the VM allocation level; (ii) declarative cross-layer experiment configuration; and (iii) automation features covering the entire experimental process, from the configuration up to the results visualization. We present proof-of-concept results that demonstrate the above capabilities in three distinct contexts: (i) a comparative evaluation of various network fabrics across different edge-oriented Kubernetes distributions; (ii) the automated deployment of EdgeNet, which is a complex edge cloud orchestration system; and (iii) an assessment of anomaly detection (AD) workflows tailored for edge environments.

An Open-Source Experimentation Framework for the Edge Cloud Continuum

TL;DR

The paper presents CODECO, an open-source Experimentation Framework designed to accelerate rapid, cross-layer testing of Kubernetes-based edge cloud deployments. It introduces a declarative Experiment Descriptor and a microservice architecture that orchestrates cluster provisioning, application deployment, experiment execution, and results processing across diverse edge environments. The authors demonstrate three proof-of-concept scenarios: benchmarking network plugins across edge-oriented Kubernetes distributions, automated deployment of EdgeNet, and online anomaly-detection workflows tailored for edge systems. These contributions enable reproducible, end-to-end experimentation with configurable cross-layer parameters, supporting integration with external testbeds and edge platforms. The framework advances practical evaluation of edge orchestration, networking, and AI-driven workflows in a scalable, extensible manner.

Abstract

The CODECO Experimentation Framework is an open-source solution designed for the rapid experimentation of Kubernetes-based edge cloud deployments. It adopts a microservice-based architecture and introduces innovative abstractions for (i) the holistic deployment of Kubernetes clusters and associated applications, starting from the VM allocation level; (ii) declarative cross-layer experiment configuration; and (iii) automation features covering the entire experimental process, from the configuration up to the results visualization. We present proof-of-concept results that demonstrate the above capabilities in three distinct contexts: (i) a comparative evaluation of various network fabrics across different edge-oriented Kubernetes distributions; (ii) the automated deployment of EdgeNet, which is a complex edge cloud orchestration system; and (iii) an assessment of anomaly detection (AD) workflows tailored for edge environments.
Paper Structure (11 sections, 6 figures)

This paper contains 11 sections, 6 figures.

Table of Contents

  1. Introduction
  2. Design and Implementation
  3. Experiment Descriptor
  4. Architecture
  5. Proof-of-concept Results
  6. Experimentation Setup
  7. Experimentation Scenarios
  8. Edge-oriented K8s Distributions and Network Plugins
  9. EdgeNet Deployment
  10. Anomaly Detection Workflows for the Edge
  11. Conclusions and Future work

Figures (6)

  • Figure 1: A Simple Experiment Definition File
  • Figure 2: Main architectural components and interactions.
  • Figure 3: CPU, RAM usage and throughput per CNI plugin, for the K8s distribution.
  • Figure 4: CPU, RAM usage and throughput per CNI plugin, for lightweight distributions.
  • Figure 5: EdgeNet installation and node contribution output commands.
  • ...and 1 more figures