Observability Glossary

P90

P90 is a statistical term commonly used in observability to understand the distribution of data. For example, if we measure the response times of a service and calculate the P90, it tells us that 90% of the response times are below this value. This is useful for understanding the worst-case performance scenario in a system.

If the P90 of the request latencies in your application is 300ms, it means that 90% of requests completed with a latency below 300ms, and 10% were above. With large datasets, it is recommended to measure worse case scenarios using P90 or other percentiles, as relying on the maximum latency can be misleading due to outliers.

Explore related concepts
Start resolving issues today.
Without the hassle.
This website uses cookies

We use cookies to offer you a better browsing experience, analyze site traffic and personalize content. By using our site, you consent to our use of cookies.