Improve documentation around requirements for k8s environment detection to work
We should improve the documentation to more clearly state what is required for the k8s memory and CPU metrics to work, namely how we identify a specific environment.
- The base requirement is that an
environmentlabel must be present. If users are bringing their own Prometheus servers, they can write a metric relabeling rule to populate this, if possible.
- Define your deployment in the format of
<CI_ENVIRONMENT_SLUG>with an optional additional two fields delimited by