[prometheus] Stabilize Prometheus -> OTLP: Counters and Gauges

[dashpole]

What are you trying to achieve?

Stabilize https://opentelemetry.io/docs/specs/otel/compatibility/prometheus_and_openmetrics/#counters and https://opentelemetry.io/docs/specs/otel/compatibility/prometheus_and_openmetrics/#gauges

I don't think there are any open questions.

[ArthurSens]

As mentioned in #4862, do we need to discuss when a Prometheus Gauge becomes a OTLP Gauge or a OTLP UpDownCounter?

I understand the difference between them from the OTLP point of view, which intended them to be aggregated differently in alerts/dashboarding, but in Prometheus, this distinction doesn't exist, or at least isn't explicit.

Do we have a reliable way to imply the difference from Prometheus Gauges...?

[ArthurSens]

Maybe?:

• If Prom gauge has no labels -> Gauge
• If Prom gauge has labels -> UpDownCounter

[dashpole]

I don't think there is a way for us to differentiate the two. Some existing prometheus gauges (e.g. container memory usage) would map to OTel UpDownCounters, and others would map to gauge (e.g. kube_pod_start_time). I think the best we can do is to say that they map to gauge, and let users re-map (e.g. with schema transforms?) metrics to the correct type after the fact.

[ArthurSens]

Oke, I'll just stabilize them as they are then