Since https://github.com/brancz/kubernetes-grafana/pull/115, upstream
grafana contains a non-empty config. Generate the grafana-config secret
unconditionally even if no user config is passed.

Adding a PodDisruptionBudget to prometheus-adapter ensure that at least
one replica of the adapter is always available. This make sure that even
during disruption the aggregated API is available and thus does not
impact the availability of the apiserver.

Signed-off-by: Damien Grisonnet <dgrisonn@redhat.com>

Signed-off-by: Simon Pasquier <spasquie@redhat.com>

* Avoid race condition when deploying quickstart example

The namespace and CRD creation must happen before any dependent objects
are created.  So we can put these in a separate directory (manifest/setup)
so they can be created before the other objects.

Some minor updates to the README and added a couple of scripts
for the quickstarts

Update travis script to avoid race condition

Signed-off-by: Paul Gier <pgier@redhat.com>

* simplify the example quickstart script and improve readme

Signed-off-by: Paul Gier <pgier@redhat.com>

* increase minikube memory to 6g for quickstart example

* Bumping prometheus-adapter and Grafana versions

* Bumping prometheus-adapter version in libsonnet file

* Regenerating lockfile

* Bumping Grafana version in libsonnet file

* Updating kustomization.yaml

* [Modified] Makefile to run kustomize.jsonnet
* [Modified] Moved kustomization from example.jsonnet to
examples/kustomize.jsonnet
* [Modified] kustomization file to land at root of kube-prometheus so
current users can continue to use `kubectl apply -f manifests`