The response time metric measures how long it takes for requests to be processed. It helps in identifying both the typical and worst-case performance of the system.
We display multiple statistical views of response times:
Max: The slowest request within the selected time range.
99th Percentile: The time within which 99% of requests are completed, highlighting rare but high-latency cases.
98th Percentile: The time within which 98% of requests are completed, useful for spotting slightly more frequent outliers.
50th Percentile (Median): The "middle" response time — half of the requests were faster, half were slower, showing typical user experience.
These metrics together provide a balanced view of normal performance and edge-case delays.