Improve Prometheus Metrics

[kevarr]

Is your feature request related to a problem?

I couldn't find any documentation for what metrics the "metrics-server" port exposes.

Additionally, I couldn't configure the Helm Chart to not scrape these metrics, as my scrape_configs are looking at the prometheus.io/* annotations and these cannot be removed without enabling a ServiceMonitor.

Describe the solution you'd like

• Helm Chart capability to remove prometheus.io/* annotations without enabling the ServiceMonitor
• Documentation on available controller metrics
• Metrics for reconciles (totals, errors successes), invalid configurations

Describe alternatives you've considered

N/A

[kevarr]

Related #3244

[huangm777]

Thank you for raising the issue! We will be taking a look at it, and of course we also welcome any community contributions to address the issue as well.

[kevarr]

Thanks @huangm777, ended up finding some metrics I could use, namely the aws_api ones. Still I think I would benefit from having some of these documented. When I find time I'll try to open a PR for this