@ericpaulsen you had mentioned linking to the replica & resources values in our Helm chart

is that this section of coder-observability/values.yaml?

This feels too separated from the content above about infrastructure metrics.

The idea is that you need to capture 2 kinds of metrics: infrastructure, and Coder-specific, and you want them both to be available in the same system for alerting and dashboards.

I wonder if that's because the subheadings under Observability are serving different purposes, with ### Observability key metrics providing more information about a specific aspect, and ### How to capture Coder server metrics with Prometheus showing steps.

I rearranged things in f110af3 to see if that helps the read-through, but I think that if we can add an additional method for capturing metrics and logs, we can split that out into its own H2

Does it make sense to add an example of how to stream coderd or provisionerd logs?

I worry a bit about mixing high-level advice with specific step-by-step instructions, since the step-by-step stuff could change and people might not realize they also have to update this document. I'd much prefer linking so that step-by-step instructions appears in one place for any specific task.

fair - that'd be a lot easier with reusable content (embeds/includes/partials, whatever), but as it is now, let's get this out and add relevant links around as we find them