Currently, these demos are only tested in a MacOS development environment. We do not cover the basics of how to setup a dev machine, but we give some hints and links you might find helpful.
To use Minikube on your machine, you have to install a hypervisor (We use hyperkit). You find detailed instructions about installation and quickstart on the Kubernetes site.
Caution! If you start minikube, it overwrites your Kubernetes config if you already have one! Make sure to backup ~/.kube/config before you start minikube.
To start your local minikube installation, use this command:
minikube start --vm-driver=hyperkit --memory 8192 --cpus 3
We use an Kubernetes ingress to access the Graylog UI via the web browser. To enable ingresses in your minikube, use this command:
minikube addons enable ingress
Although you can work with your Kubernetes cluster purely via the command line (kubectl), sometimes you might find a UI useful. To start a browser-based Dashboard UI:
minikube dashboard
As described in the official docs, ElasticSearch needs some special settings for virtual memory in the machine that it is running in.
One way to solve this is to ssh into the running minikube and set the value:
minikube ssh
when you're in the minikube shell:
sudo sysctl -w vm.max_map_count=262144
Unfortunately, you have to set this after every restart of the minikube.
Alternatively you can achieve this with an init container as described here.