The impact of using eBPF technology on the performance of networking solutions in a Kubernetes cluster
Article Sidebar
Open full text
Issue Vol. 35 (2025)
-
Analysis of the impact of selected user interface elements on its usability
113-120
-
Performance comparison of development frameworks in selected environments in REST API architecture
121-128
-
Comparison of data processing efficiency in Java and Scala
129-136
-
Comparative analysis of Python and Rust: evaluating their combined impact on performance
137-141
-
Comparison of the flexibility of standard Salesforce components and custom components in Lightning Web Components
142-149
-
The impact of using eBPF technology on the performance of networking solutions in a Kubernetes cluster
150-158
-
A comparative analysis of web application test automation tools
159-165
-
Advancing Bangla typography: machine learning and transfer learning based font detection and classification approach using the ‘Bang-laFont45’ dataset
ML and TL based font detection and classification approach using ‘Bang-laFont45’ dataset166-174 -
Performance comparison of CRUD operations in Spring Boot and ASP.NET Core frameworks
175-183
-
The impact of relational and non-relational databases on application performance
184-190
-
An analysis of the quality of interfaces of selected universities in Lublin
191-196
-
Comparative analysis of the performance of Unity and Unreal Engine
197-201
-
Analysis the efficiency of object detection in images using machine learning libraries in Python
202-208
-
Comparative analysis of database access performance of the Hibernate framework and the Jooq library
209-215
-
Security vulnerabilities in C++ programs
216-223
-
Impact of web application universal design on accessibility and usability
224-231
Main Article Content
DOI
Authors
Abstract
The aim of this study was to investigate the impact of eBPF technology on the performance of network solutions in Kubernetes clusters. Two configurations were compared: a traditional iptables-based setup and eBPF based solution via the Cilium networking plugin. Performance tests were conducted, measuring throughput, latency, CPU usage, and memory consumption under unloaded and loaded conditions. The results indicate that the traditional configuration achieved higher throughput and lower latency in unloaded scenarios. However, under load, the eBPF-enabled cluster demonstrated advantages, including reduced CPU and memory usage and slightly improved latency. This study highlights the potential of eBPF as an efficient technology for Kubernetes environments, particularly in scenarios demanding high performance and resource efficiency.
