Question

A cloud operations team monitors a customer-facing API. The dashboard shows the average (mean) response time is a healthy 120 ms, and the team receives no alerts. However, the support desk is flooded with complaints from users experiencing multi-second delays. Which monitoring change would BEST reveal the true impact on affected users?

Accepted Answer

Track and alert on the 95th and 99th percentile (p95/p99) latency metrics instead of relying on the mean

Answer

Increase the sampling interval so metrics are collected every 60 seconds rather than every 10 seconds

Answer

Raise the mean latency alert threshold from 120 ms to 500 ms to reduce false negatives

Answer

Switch the dashboard to display total request count per minute instead of response time