I recently had a situation in which the Prometheus console displayed no data, despite the Fuse applications it was supposed to monitor exposing metrics on their Prometheus ports. It was impossible to troubleshoot the problem, because the Prometheus console pod did not log any useful diagnostic data. It just looked fine; but clearly it wasn't.

What we need is a configurable and well-documented way to increase the log verbosity of the Prometheus console, ideally to the point at which we can see how it is interpreting its configuration and we can inspect the specific interactions it makes with the pods it is supposed to monitor.