At the moment, the starting of k8s and of the application is made in the same script.
This is fine for Vagrant provisionning where everything is run as a single script since the VM is rather empty (well, it has docker).
However, testing other ways of running k8s isn't facilitated. For instance I've tested running k8s with KinD (Kubernetes in Docker), with handles the starting of the cluster.
I'd then prefer if the kubernetes start and the launch of syringe + antidote + ingress are separated in different script.
Having 2 provisioning script shouldn't be blocking for Vagrant, whereas one could only run the last one separately when using minikube, kind or other ways to start Antidote over an existing k8s cluster.
At the moment, the starting of k8s and of the application is made in the same script.
This is fine for Vagrant provisionning where everything is run as a single script since the VM is rather empty (well, it has docker).
However, testing other ways of running k8s isn't facilitated. For instance I've tested running k8s with KinD (Kubernetes in Docker), with handles the starting of the cluster.
I'd then prefer if the kubernetes start and the launch of syringe + antidote + ingress are separated in different script.
Having 2 provisioning script shouldn't be blocking for Vagrant, whereas one could only run the last one separately when using minikube, kind or other ways to start Antidote over an existing k8s cluster.
Hope this makes sense.