Add FAQ page for OCP monitoring

[simonpasquier]

List of things that we are often asked about on Slack, emails, ... which would need to be answered in the FAQ.

Feel free to add any suggestion in the comments :)

[simonpasquier]

• OCP support
  • Does Red Hat support Prometheus remote write to <insert 3rd party system>?
    • Red Hat supports only the remote write sender side (e.g. Prometheus). Issues on the receiving side need to be supported by the customer.
• OCP developers
  • How do I push metrics to telemetry?
    • Link to existing doc
  • How do I delete a label value when it stops existing?
    • Link to the custom collector pattern

[simonpasquier]

Why doesn't OCP monitoring support <tweaking the scrape interval/deploying custom exporters/...>?

• OCP monitoring is primarily meant for monitoring the platform components. By controlling the configuration options, we limit the risk of users DoS-ing the monitoring stack.

[simonpasquier]

What are the API endpoints available? What are the required RBAC roles/cluster roles for each?

[simonpasquier]

Explain how user-defined monitoring works in practice (namespace enforcement for metrics & rules).

[sthaha]

Why was X feature added or removed?

We often get questions of the form

Release 4.Y.Z removes/adds feature , what is the reason behind that?

We need a way to trace from Release Notes to Epic.

[sthaha]

Explain why running another PO isn't a supported usecase.

[sthaha]

Explain the use/abuse of "openshift.io/cluster-monitoring: "true".

• Is it supported ?
• What is the intended usecase ?

[sthaha]

What is the supported way for installing Grafana and configuring to incluster prometheus as datasource?

[ssonigra]

Q. How to list all the available metrics in OCP

A. https://access.redhat.com/solutions/6984824