TY - GEN
T1 - A Comparative Analysis of Self-Hosted Nomad and Managed Kubernetes for Container Orchestration
AU - Cardas, Alexandru
AU - Ayala-Rivera, Vanessa
AU - Portillo-Dominguez, A. Omar
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Containerization has become central to scalable application deployment, driving the rise of microservice architectures. While these architectures enhance modularity and scalability, they introduce operational complexity that necessitates robust orchestration solutions. However, there is limited empirical research comparing widely used orchestrators (particularly managed Kubernetes services and alternatives like Nomad) in real-world cloud environments. This paper addresses this gap by evaluating the performance, cost efficiency, and scalability of AWS Elastic Kubernetes Service (EKS) versus a self-hosted Nomad deployment on AWS, using controlled experiments that simulate production conditions. Our experimental results provide practical guidance for choosing between managed and self-hosted orchestration solutions based on workload demands and operational constraints (e.g., our results showed that Nomad achieves up to 20% lower request latency and 3.6× lower infrastructure costs compared to EKS in small-to-medium scale deployments).
AB - Containerization has become central to scalable application deployment, driving the rise of microservice architectures. While these architectures enhance modularity and scalability, they introduce operational complexity that necessitates robust orchestration solutions. However, there is limited empirical research comparing widely used orchestrators (particularly managed Kubernetes services and alternatives like Nomad) in real-world cloud environments. This paper addresses this gap by evaluating the performance, cost efficiency, and scalability of AWS Elastic Kubernetes Service (EKS) versus a self-hosted Nomad deployment on AWS, using controlled experiments that simulate production conditions. Our experimental results provide practical guidance for choosing between managed and self-hosted orchestration solutions based on workload demands and operational constraints (e.g., our results showed that Nomad achieves up to 20% lower request latency and 3.6× lower infrastructure costs compared to EKS in small-to-medium scale deployments).
KW - Cloud Computing
KW - Container Orchestration
KW - Kubernetes (EKS)
KW - Nomad
KW - Performance Engineering
UR - https://www.scopus.com/pages/publications/105024940291
U2 - 10.1109/ICITCE65255.2025.11210789
DO - 10.1109/ICITCE65255.2025.11210789
M3 - Conference contribution
AN - SCOPUS:105024940291
T3 - IEEE International Conference on Industrial Technology and Computer Engineering, ICITCE 2025
SP - 5
EP - 10
BT - IEEE International Conference on Industrial Technology and Computer Engineering, ICITCE 2025
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2025 IEEE International Conference on Industrial Technology and Computer Engineering, ICITCE 2025
Y2 - 13 September 2025 through 14 September 2025
ER -