Skip to content
All concepts

Concept

p99 latency

p99 latency is the response time that 99 out of 100 requests beat. It describes the slow tail that an average or a median hides. Because one page usually makes many calls, a large share of users meet the tail even when only one call in a hundred is slow.

Watch p99 sit next to the median: it shows the slow calls that the median never moves for. A live step from the lesson p50 vs p99 (Observability). Timings and counts are illustrative.

Where you'll see it

Sources: The Tail at Scale (Dean & Barroso, Google, CACM 2013) ↗ · Prometheus docs: Histograms and summaries ↗ · Report a problem

The definition is in our own words; the sources are those of the lesson p50 vs p99.