This sets up a public network (you'll get prompted for which of your NICs should be used). This emulates a public-routable VM. Take precautions (tho testing the precautions is somewhat the point here)
Once up and running,
Port mappings:
| From | To | |
|---|---|---|
| Bridge IP : 80,443 | -> | Worker Node : 80,443 |
| Bridge IP : 30000-32767 | -> | Worker Node : 30000-32767 |
| Bridge IP : 6443 | -> | Master Node : 6443 |
| Bridge IP : 2375,5000 | -> | VM Docker Daemon / Registry |
After vagrant up has completed, there will be a full kube running inside the VM with 5 'machines' running (each machine is an lxc container), providing easyrsa, etcd, flannel, kubernetes-master and a single kubernetes-worker.
A config file (called kubeconfig) to let you talk to the kube will be generated into the same directory as the Vagrantfile.
Test access to your cluster using kubectl --kubeconfig=./kubeconfig cluster-info
To work with custom images, the VM also exposes a docker daemon on port 2375 (the default), you can use this by setting the env var DOCKER_HOST to the Bridge IP. The VM will tell you the ip and port when it has finished initialising.
eg. export DOCKER_HOST=192.168.1.100
To push images so they can be used within the kube instances, tag the image with the repository name, then push the image to the repository.
eg. docker tag your-image:yourtag 192.168.1.100:5000/your-image:yourtag
then docker push 192.168.1.100:5000/your-image:yourtag
Now you can refer to the images from Kube yaml using the image name 192.168.1.100:5000/your-image:yourtag