> The irony is that this may be a $0 revenue user for Grafana Labs.

Why is that ironic? Since Mimir is open-source, $0 revenue users are expected. AFAIK, Grafana Labs relies heavily on go, typescript, and linux, without necessarily being their top financial contributor. They could have kept Mimir proprietary like Splunk, but whether that would have attracted the same level of adoption or community contribution is another matter.

Directly emitting metrics using OTLP instead of having the OTel receiver scrape the metrics endpoint is interesting. I never made that move because the Prometheus metrics endpoint works and is so simple, and it's what most projects (eg kubernetes) use.

Curious why the team choose Grafana Mirmir over VM cluster?

> given Prometheus’s widespread adoption and proven reliability in diverse environments.

I have used Prometheus a lot. Reliable is not a word I would associate with it.