A simple load generator for kubernetes using helm.
There are times when you just want to generate some background load to stress-test your software on a cluster. This is especially useful when you're looking to test the scale out and in properties of your cluster as you deploy and load-test applications.
This chart is really quite simple. It uses a short ruby script in a configmap that gets mounted and run on a ruby-alpine image. The current defaults will genereate about a 1000 milli-cpu load on an m3-large EC2 instance, so simply multiplying the number of replicas will map pretty evenly to CPUs.
To Install
-
Setup your KUBECONFIG
-
Clone this repository
-
Run the following
helm install load-gen --name load-gen --set replicaCount={how many cpus}