small golang service to generate cpu load. and a container with the load generator hey. the service is configured to autoscale from 1-50 pods based on pod load of 50% .
this is done for testing of different horizontal pod autoscaler algoritms. for more load just change the hey -c concurrent connection flag, -n total number of requests and the replica count of the hey deployment.
This repo is works with the standard go tools. It also works with Bazel.
Building with Bazel requires a bazel 5.3.0+ installation and running the following command:
bazel build //...
After the build is complete the k8s-load binary lives under the bazel-bin directory:
bazel-bin/k8s-load
kubectl apply -f deployment-hey.yaml
kubectl apply -f deployment-service.yaml