Tipo di tesi
Tesi di laurea magistrale
Titolo
Platform Selection in Network Emulation: An Empirical Comparison of EVE-NG and Containerlab.
Corso di studi
INFORMATICA E NETWORKING
Parole chiave
- Containerlab
- EVE-NG
- Network Emulation
- Performance Evaluation
- Resilience
- Virtualization
Data inizio appello
29/05/2026
Consultabilità
Non consultabile
Data di rilascio
29/05/2029
Riassunto (Inglese)
Network emulation platforms are widely used to validate network architectures and protocols before real-world deployment. Among the available solutions, EVE-NG (Emulated Virtual Environment – Next Generation) and Containerlab support comparable multi-node topologies while relying on fundamentally different virtualization approaches, namely virtual machines and container-based virtualization. Despite their increasing adoption in both academic and industrial environments, limited empirical evidence is available regarding their comparative performance, resource efficiency, and resilience under equivalent experimental conditions.
This thesis presents a systematic comparison between EVE-NG and Containerlab by deploying equivalent network topologies with identical addressing schemes, routing configurations, and traffic generation workloads. The experimental campaign evaluates key performance indicators, including throughput, latency, CPU utilization, and memory consumption, together with resilience metrics obtained through controlled failure injection scenarios involving firewall crashes, link disruptions, routing protocol failures, and CPU saturation events.
The obtained results show that, within the considered experimental setup, Containerlab generally achieves higher throughput, lower latency, and improved resource efficiency compared with EVE-NG, which exhibits higher virtualization overhead and increased resource consumption. Furthermore, Containerlab demonstrates faster recovery in most failure scenarios, whereas EVE-NG achieves slightly shorter recovery times only during link restoration experiments.
Overall, the study demonstrates that the choice of the emulation platform can significantly affect the observed networking behavior and experimental outcomes. The presented findings provide practical insights for researchers, educators, and network engineers involved in the design and evaluation of virtualized networking environments.
