Listen for PrometheusRule resources and configure in Grafana Mimir/Cortex

[flokli]

It'd be really awesome if GrafanaAgent could listen for PrometheusRule resources, and create remote Mimir/Cortex configuration.

While it seems it's possible to configure these manually via Grafana, a lot of applications ship PrometheusRule resources to describe aggregations or alerts.

Using this would mean there's no need to deploy alertmanager in a downstream cluster, and do aggregation there, but it could all live in the monitoring system.

[rfratto]

I'd love this! This is somewhat similar to #523, which we'd probably add first before supporting PrometheusRule resources.

[TarekAS]

This would be great to have if we want to use Mimir as drop-in replacement for kube-prometheus-stack

[jhohertz]

#1986 seems to cover the same subject, appears to have some attention.