1. Proposed title of this feature request

Deploy and enable ironic-prometheus-exporter as part of the OpenShift IPI installation on baremetal

2. What is the nature and description of the request?

https://github.com/openstack/ironic-prometheus-exporter provides a way to expose metrics from baremetal hosts so prometheus can scrape them. Currently ironic-prometheus-exporter is already included in the metal3-ironic-conductor container which is part of the metal3 pod, but is disabled by default. Also, it requires another container to serve those collected metrics to a prometheus endpoint, a service and a servicemonitor.
It would be nice to have a flag/tweak/something (maybe just a configmap?) so users can enable/disable the metrics collection easily. As a bonus, maybe add some grafana dashboards to show power consumption, temperatures, etc.

3. Why does the customer need this? (List the business requirements here)

Usually customers monitor their bare metal hosts using proprietary tools created by the hardware manufacturer, but most of them (all of them?) aren't containerized nor run on k8s/OCP, so having something that is independent of the hardware vendor is helpful (at the end of the day, i-p-e perform standard queries to redfish/ipmi/etc. so it can be easily extended)

4. List any affected packages or components.

ironic-prometheus-agent is already packaged but it is missing an easy way to use it currently in OpenShift