[DOC] Monitoring metrics are misnamed

[agardnerIT]

Change in the documentation

This page shows:

flagd expose following metrics,

http.server.duration
http.server.response.size
http.server.active_requests
feature_flag.flagd.impression
feature_flag.flagd.evaluation.reason

1) Typo: expose should be exposes 2) The metrics are misnamed:

• http.server.duration is actually # TYPE http_server_duration_seconds histogram
• http.server.response.size is actually # TYPE http_server_response_size_bytes histogram
• http.server.active_requests is actually # TYPE http_server_active_requests gauge
• feature_flag.flagd.impression is actually # TYPE feature_flag_flagd_impression_total counter
• feature_flag.flagd.evaluation.reason is actually # TYPE feature_flag_flagd_evaluation_reason_total counter 3) We should consider adding a line to this page to say "These metrics will only appear after the first flag evaluation"

[beeme1mr]

Hey @agardnerIT, I've fixed the typos. The documented metric names are correct and follow OTel's recommended naming convention. However, Prometheus, Dynatrace, and likely others have different metric naming requirements. I've added a disclaimer that the metric names may be transformed by the consuming monitoring tool.