Skip to content

Simple wrapper around iperf3 to measure network bandwidth from all nodes of a Kubernetes cluster

License

Notifications You must be signed in to change notification settings

Pharb/kubernetes-iperf3

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

kubernetes-iperf3

Simple wrapper around iperf3 to measure network bandwidth from all nodes of a Kubernetes cluster.

How to use

Make sure you are using the correct cluster context before running this script: kubectl config current-context

$ ./iperf3.sh

Any options supported by iperf3 can be added, e.g.:

$ ./iperf3.sh -t 2

NetworkPolicies

If you need NetworkPolicies you can install it:

$ kubectl apply -f network-policy.yaml

And cleanup afterwards:

$ kubectl delete -f network-policy.yaml

Output:

deployment.apps "iperf3-server-deployment" created
service "iperf3-server" created
daemonset.apps "iperf3-clients" created
Waiting for iperf3 server to start...
Waiting for iperf3 server to start...
Waiting for iperf3 server to start...
Server is running

Client on 172.20.47.197:  Connecting to host iperf3-server, port 5201
Client on 172.20.47.197:  [  4] local 100.96.0.28 port 37580 connected to 100.65.13.40 port 5201
Client on 172.20.47.197:  [ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
Client on 172.20.47.197:  [  4]   0.00-1.00   sec  3.59 GBytes  30.8 Gbits/sec    0   4.00 MBytes
Client on 172.20.47.197:  [  4]   1.00-2.00   sec  3.66 GBytes  31.5 Gbits/sec    0   4.00 MBytes
Client on 172.20.47.197:  - - - - - - - - - - - - - - - - - - - - - - - - -
Client on 172.20.47.197:  [ ID] Interval           Transfer     Bandwidth       Retr
Client on 172.20.47.197:  [  4]   0.00-2.00   sec  7.25 GBytes  31.1 Gbits/sec    0             sender
Client on 172.20.47.197:  [  4]   0.00-2.00   sec  7.25 GBytes  31.1 Gbits/sec                  receiver
Client on 172.20.47.197:
Client on 172.20.47.197:  iperf Done.

Client on 172.20.57.129:  Connecting to host iperf3-server, port 5201
Client on 172.20.57.129:  [  4] local 100.96.1.31 port 55166 connected to 100.65.13.40 port 5201
Client on 172.20.57.129:  [ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
Client on 172.20.57.129:  [  4]   0.00-1.00   sec   954 MBytes  8.00 Gbits/sec    5   3.10 MBytes
Client on 172.20.57.129:  [  4]   1.00-2.00   sec   948 MBytes  7.95 Gbits/sec    1   4.22 MBytes
Client on 172.20.57.129:  - - - - - - - - - - - - - - - - - - - - - - - - -
Client on 172.20.57.129:  [ ID] Interval           Transfer     Bandwidth       Retr
Client on 172.20.57.129:  [  4]   0.00-2.00   sec  1.86 GBytes  7.97 Gbits/sec    6             sender
Client on 172.20.57.129:  [  4]   0.00-2.00   sec  1.85 GBytes  7.93 Gbits/sec                  receiver
Client on 172.20.57.129:
Client on 172.20.57.129:  iperf Done.

deployment.apps "iperf3-server-deployment" deleted
service "iperf3-server" deleted
daemonset.apps "iperf3-clients" deleted

Caution

This script could potentially disrupt normal cluster operations by causing high CPU and network load. Use with care.

How it works

The script will run an iperf3 client inside a pod on every cluster node including the Kubernetes master. Each iperf3 client will then sequentially run the same benchmark against the iperf3 server running on the Kubernetes master.

All required Kubernetes resources will be created and removed after the benchmark finished successfully.

This has been tested with v1.9.6, v1.10.3 and v1.11.6 of Kubernetes.

The latest version of this Docker image is used to run iperf3: https://hub.docker.com/r/networkstatic/iperf3/

Details on how to use iperf3 can be found here: https://github.com/esnet/iperf

Thanks

Thanks to the iperf maintainers for providing such a great tool.

License

MIT

About

Simple wrapper around iperf3 to measure network bandwidth from all nodes of a Kubernetes cluster

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages