A publicly accessible web server receives 1,000 requests per hour and has an average error rate of 1–3%.
What information would you store in Prometheus?
Prometheus stores numeric time-series metrics suitable for aggregation, alerting, and rate calculations. Request and error counters provide request and error rates, while histograms or summaries capture request durations. Detailed logs and stack traces are high-cardinality event data and belong in a logging or tracing system rather than Prometheus. Metric types | Prometheus
Community Discussion