Efficient Container Image Distribution for Multi-Tenant Kubernetes Clusters

The distributed Kubernetes environment utilizes container images to run designated workloads. Image registries, which store these container images, can be located either inside or outside the Kubernetes cluster, depending on the product and application requirements. Container image sizes can vary according to the type and scale of workload. Consequently, the application startup time depends on the duration of the image download phase. If the image registry is hosted on a low-bandwidth link outside the cluster, this can negatively impact the application's startup time. Prolonged download times may saturate the internet link, potentially violating SLA agreements, and causing issues in production environments. Previous approaches have primarily focused on building a P2P network across Kubernetes nodes to facilitate the local distribution of images. However, in multi-tenant environments or when disk space is constrained, this method becomes impractical. In this paper, we introduce a novel method that distributes container images only to the nodes when required, leveraging the local high-bandwidth network. This approach reduces disk space utilization and decreases the application's startup time. Under resource-intensive conditions, the proposed method achieved a reduction in image download time of approximately 50 %.

Authors

Institutions

Publication Details

Published
2025-06-02
DOI
https://doi.org/10.1109/mipro65660.2025.11131820
Primary Topic
Medical Image Segmentation Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Efficient Container Image Distribution for Multi-Tenant Kubernetes Clusters

Dario Sindičić, Norman Momčilović
Medical Image Segmentation Techniques
article

Efficient Container Image Distribution for Multi-Tenant Kubernetes Clusters

Dario Sindičić, Norman Momčilović
article en

Abstract

The distributed Kubernetes environment utilizes container images to run designated workloads. Image registries, which store these container images, can be located either inside or outside the Kubernetes cluster, depending on the product and application requirements. Container image sizes can vary according to the type and scale of workload. Consequently, the application startup time depends on the duration of the image download phase. If the image registry is hosted on a low-bandwidth link outside the cluster, this can negatively impact the application's startup time. Prolonged download times may saturate the internet link, potentially violating SLA agreements, and causing issues in production environments. Previous approaches have primarily focused on building a P2P network across Kubernetes nodes to facilitate the local distribution of images. However, in multi-tenant environments or when disk space is constrained, this method becomes impractical. In this paper, we introduce a novel method that distributes container images only to the nodes when required, leveraging the local high-bandwidth network. This approach reduces disk space utilization and decreases the application's startup time. Under resource-intensive conditions, the proposed method achieved a reduction in image download time of approximately 50 %.

University of Zagreb (HR)
Openalex Percentile: Top 78%
Medical Image Segmentation Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.