KUBETWIN

It enables what-if analysis by simulating IT services deployed on Kubernetes running in Compute Continuum-like scenarios.

Explore on GitHub

Kubernetes is today’s de facto standard for deploying and managing IT services, powering everything from AWS, Azure, and Google Cloud to edge and IoT infrastructures.

Yet configuring these systems is complex and time-consuming.

KubeTwin simplifies this process by providing a safe, software-only environment where engineers can test configurations, anticipate issues, and achieve optimal performance.

Companies like Netflix use Chaos Engineering — deliberately injecting faults in production systems to test their resilience.

While effective, this method is risky and costly.

KubeTwin offers the same insights at a fraction of the risk: experiments happen on the digital twin, not on live systems.

Faults, workloads, and recovery strategies can be explored safely and efficiently.

KubeTwin is a collaborative effort involving: 

Big Data & Compute Continuum Lab, University of Ferrara, Italy

University of Bologna, Italy

St. John’s University, NY, USA

IBM Consulting, NY, USA

Ghent University, Belgium

KubeTwin's current publications list:

D. Borsatti et al., "Modeling Digital Twins of Kubernetes-Based Applications," 2023 IEEE Symposium on Computers and Communications (ISCC), Gammarth, Tunisia, 2023, pp. 219-224, doi: 10.1109/ISCC58397.2023.10217853. 

L. Manca et al., "Characterization of Microservice Response Time in Kubernetes: A Mixture Density Network Approach," 2023 19th International Conference on Network and Service Management (CNSM), Niagara Falls, ON, Canada, 2023, pp. 1-9, doi: 10.23919/CNSM59352.2023.10327842. 

D. Borsatti et al., "KubeTwin: A Digital Twin Framework for Kubernetes Deployments at Scale," in IEEE Transactions on Network and Service Management, vol. 21, no. 4, pp. 3889-3903, Aug. 2024, doi: 10.1109/TNSM.2024.3405175. 

M. Zaccarini et al., "TELKA: Twin-Enhanced Learning for Kubernetes Applications," 2024 IEEE Symposium on Computers and Communications (ISCC), Paris, France, 2024, pp. 1-6, doi: 10.1109/ISCC61673.2024.10733736. 

M. Zaccarini et al., "Chaos Engineering Based Kubernetes Pod Rescheduling Through Deep Sets and Reinforcement Learning," NOMS 2025-2025 IEEE Network Operations and Management Symposium, Honolulu, HI, USA, 2025, pp. 1-7, doi: 10.1109/NOMS57970.2025.11073590. 

KubeTwin has received outstanding recognition across the Network and Service Management research community:

2024 CNOM Best Paper Award 

selected by the IEEE ComSoc Technical Committee on Network Operation and Management (CNOM)

2023 IFIP-supported Best Paper Award 

at the 19th International Conference on Network and Service Management 

(CNSM 2023), Niagara Falls, Canada

Awarded Paper: Characterization of Microservice Response Time in Kubernetes: A Mixture Density Network Approach

DOI: 10.23919/cnsm59352.2023.10327842

Authors: 

Lorenzo Manca University of Bologna, Bologna, Italy

Davide Borsatti University of Bologna, Bologna, Italy

Filippo Poltronieri Big Data & Compute Continuum Lab, University of Ferrara, Ferrara, Italy

Mattia Zaccarini Big Data & Compute Continuum Lab, University of Ferrara, Ferrara, Italy

Domenico Scotece University of Bologna, Bologna, Italy

Gianluca Davoli University of Bologna, Bologna, Italy

Luca Foschini University of Bologna, Bologna, Italy

Genady Ya. Grabarnik Department of Mathematics and Computer Science, St. John's University, NY, USA

Larisa Shwartz HCS (Hybrid-Cloud Services) Department, IBM Consulting, NY, USA

Cesare Stefanelli Distributed Systems Research Group, University of Ferrara, Ferrara, Italy

Mauro Tortonesi Big Data & Compute Continuum Lab, University of Ferrara, Ferrara, Italy

Walter Cerroni University of Bologna, Bologna, Italy

Digital Twin · Kubernetes · Simulation · Configuration Optimization · Microservices · Orchestration · Troubleshooting · Chaos Engineering · Resilience · Scalability

GitHub Repository: github.com/DSG-UniFE/KubeTwin 

Coming Soon: Official documentation and demo portal.