Access Logs & Metrics

Two pieces of observability every server needs:

Access log — one line per request:

192.168.1.5 - - [17/Mar/2025:09:15:23 +0000] "GET /api/users HTTP/1.1" 200 482 "-" "curl/8.5.0" 23ms

This is the Common Log Format (Apache) extended with response time and user agent. Tools like AWStats, GoAccess, Splunk parse it for free.

Fields:

Metrics (Prometheus / OpenTelemetry):

• http_requests_total{method, status, route} — counter
• http_request_duration_seconds_bucket{...} — histogram
• http_request_size_bytes / http_response_size_bytes — summary

Aggregating these into dashboards is what tells you "p99 of /api/users got slow" before users complain.