On the Stability of the Kubernetes Horizontal Autoscaler Control Loop

Kubernetes is a widely used platform for deploying and managing containerized applications due to its efficient elastic capabilities. The Horizontal Pod Autoscaler (HPA) in Kubernetes independently adjusts the number of pods for each service, yet these services often operate in an interconnected man...

Full description

Saved in:
Bibliographic Details
Main Authors: Berta Serracanta, Andor Lukacs, Alberto Rodriguez-Natal, Albert Cabellos, Gabor Retvari
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10829852/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850094169452707840
author Berta Serracanta
Andor Lukacs
Alberto Rodriguez-Natal
Albert Cabellos
Gabor Retvari
author_facet Berta Serracanta
Andor Lukacs
Alberto Rodriguez-Natal
Albert Cabellos
Gabor Retvari
author_sort Berta Serracanta
collection DOAJ
description Kubernetes is a widely used platform for deploying and managing containerized applications due to its efficient elastic capabilities. The Horizontal Pod Autoscaler (HPA) in Kubernetes independently adjusts the number of pods for each service, yet these services often operate in an interconnected manner. This study aims to understand the effects of autoscaling events on a graph of interconnected services. To achieve this, we apply control theory to model the HPA’s behavior. We analyze the stability of this model, perform numerical simulations, and deploy a real testbed to evaluate the performance. Our findings demonstrate that the control theory-based model accurately predicts the HPA’s behavior, ensuring system stability with CPU utilization meeting desired thresholds and no traffic loss after a transitional period. The model provides insights into optimizing resource scheduling and improving application performance in Kubernetes environments. Additionally, we extend our model to the whole service graph to understand how individual scaling decisions influence the complex graphs of cloud applications.
format Article
id doaj-art-5550fbb44f13438687e6958ecc076db4
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-5550fbb44f13438687e6958ecc076db42025-08-20T02:41:43ZengIEEEIEEE Access2169-35362025-01-01137160716610.1109/ACCESS.2025.352675110829852On the Stability of the Kubernetes Horizontal Autoscaler Control LoopBerta Serracanta0https://orcid.org/0000-0003-3195-2576Andor Lukacs1https://orcid.org/0000-0003-0043-1591Alberto Rodriguez-Natal2https://orcid.org/0000-0002-4239-5309Albert Cabellos3Gabor Retvari4https://orcid.org/0000-0002-5958-7817Department of Computer Architecture, Universitat Politècnica de Catalunya, Barcelona, SpainFaculty of Mathematics and Computer Science, Babeş-Bolyai University, Cluj-Napoca, RomaniaCisco, Madrid, SpainDepartment of Computer Architecture, Universitat Politècnica de Catalunya, Barcelona, SpainDepartment of Telecommunications and Artificial Intelligence, Budapest University of Technology and Economics, Budapest, HungaryKubernetes is a widely used platform for deploying and managing containerized applications due to its efficient elastic capabilities. The Horizontal Pod Autoscaler (HPA) in Kubernetes independently adjusts the number of pods for each service, yet these services often operate in an interconnected manner. This study aims to understand the effects of autoscaling events on a graph of interconnected services. To achieve this, we apply control theory to model the HPA’s behavior. We analyze the stability of this model, perform numerical simulations, and deploy a real testbed to evaluate the performance. Our findings demonstrate that the control theory-based model accurately predicts the HPA’s behavior, ensuring system stability with CPU utilization meeting desired thresholds and no traffic loss after a transitional period. The model provides insights into optimizing resource scheduling and improving application performance in Kubernetes environments. Additionally, we extend our model to the whole service graph to understand how individual scaling decisions influence the complex graphs of cloud applications.https://ieeexplore.ieee.org/document/10829852/Cloud autoscalingcontrol theoryHorizontal Pod AutoscalerKubernetesmicroservices architecturenumerical simulations
spellingShingle Berta Serracanta
Andor Lukacs
Alberto Rodriguez-Natal
Albert Cabellos
Gabor Retvari
On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
IEEE Access
Cloud autoscaling
control theory
Horizontal Pod Autoscaler
Kubernetes
microservices architecture
numerical simulations
title On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
title_full On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
title_fullStr On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
title_full_unstemmed On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
title_short On the Stability of the Kubernetes Horizontal Autoscaler Control Loop
title_sort on the stability of the kubernetes horizontal autoscaler control loop
topic Cloud autoscaling
control theory
Horizontal Pod Autoscaler
Kubernetes
microservices architecture
numerical simulations
url https://ieeexplore.ieee.org/document/10829852/
work_keys_str_mv AT bertaserracanta onthestabilityofthekuberneteshorizontalautoscalercontrolloop
AT andorlukacs onthestabilityofthekuberneteshorizontalautoscalercontrolloop
AT albertorodrigueznatal onthestabilityofthekuberneteshorizontalautoscalercontrolloop
AT albertcabellos onthestabilityofthekuberneteshorizontalautoscalercontrolloop
AT gaborretvari onthestabilityofthekuberneteshorizontalautoscalercontrolloop