Extend support for Prometheus exporters

As is now, the chart only accounts for puppetdb-exporter and it does not allow the user to utilize any metrics scraping implementation other than the one provided by the use of monitoring.coreos.com/v1 which is not always the case.

The changes in this PR allow users to choose whether they want a serviceMonitor deployed or not on the account that they might have Prometheus scraping all services/pods/etc via the Kubernetes API.

In addition to that, it adds jmx-exporter which is commonly used to get all of the Puppetserver mbean metrics in the Prometheus format allowing for insightful dashboards/alerts. These exporters are deployed next to each Puppetserver, be it master or compiler. It is up to the user to ensure that the Puppetserver exposes JMX - I've added a comment in the extraEnv that contains an example of it.

Do note that this introduces a breaking change for current users of the metrics: section in values.yaml due to support of multiple exporters, each exporter is nested under an object representing it.

This PR also adds snapshots for all unittests that make it easier catch unexpected changes when editing templates. Note, that this will require updating snapshots when the version is incremented in Chart.yaml too - this is done by running helm unittest ./ -u since the version number is contained within them in all of the resources.

Removing the commented out prometheusRule from values.yaml as it is not really used anywhere even if you were to uncomment it (from what I can tell).

puppetlabs / puppetserver-helm-chart

Extend support for Prometheus exporters #200