k8s-netperf - Kubernetes Network Performance

May 8, 2026 ยท View on GitHub

k8s-netperf Logo

k8s-netperf - Kubernetes Network Performance

A tool to run network performance tests in Kubernetes clusters.

Overview

k8s-netperf is a comprehensive network performance testing tool designed specifically for Kubernetes environments. It supports multiple benchmark tools (netperf, iperf3, uperf and ib_write_bw) and can test various network scenarios including pod-to-pod, host networking, and cross-availability zone communications.

Supported Benchmarks

ToolTestStatus
netperfTCP_STREAMWorking
netperfUDP_STREAMWorking
netperfTCP_RRWorking
netperfUDP_RRWorking
netperfTCP_CRRWorking
uperfTCP_STREAMWorking
uperfUDP_STREAMWorking
uperfTCP_RRWorking
uperfUDP_RRWorking
iperf3TCP_STREAMWorking
iperf3UDP_STREAMWorking
ib_write_bwUDP_STREAMWorking

Quick Start

Install latest stable version with:

curl -Ls https://raw.githubusercontent.com/cloud-bulldozer/k8s-netperf/refs/heads/main/hack/install.sh | sh

Note

Default installation path is ${HOME}/.local/bin/, you can change it by setting the INSTALL_DIR environment variable to the desired path. before running the script

Then create the required resources:

$ kubectl create ns netperf
$ kubectl create sa netperf -n netperf
$ k8s-netperf --config netperf.yml

Documentation

For detailed documentation, please refer to the following guides:

  • Setup and Installation - How to build, install, and configure k8s-netperf
  • Configuration - Configuration file formats, benchmark options, and OpenSearch integration
  • Advanced Usage - VMs, User Defined Networks (UDN), bridge networks, SR-IOV, MACVLAN and external servers
  • Output and Results - Understanding test output, pass/fail criteria, and CSV exports

Features

  • Multiple benchmark tools: Support for netperf, iperf3, uperf and ib_write_bw
  • Flexible test scenarios: Pod-to-pod, host networking, cross-AZ testing
  • Virtual Machine support: Test with KubeVirt VMs
  • Advanced networking: User Defined Networks (UDN), bridge interfaces, SR-IOV and MACVLAN
  • Pass/fail validation: Automated performance regression detection
  • Result archiving: CSV export and OpenSearch integration
  • Prometheus integration: System metrics collection during tests

Contributing

Please see CONTRIBUTING.md for guidelines on how to contribute to this project.