Latency
The time elapsed between a request being made and a response being received.
Definition
Latency is the time it takes for a system to respond to a request — measured from the moment the request is sent to the moment the response is received. It is one of the four golden signals of site reliability engineering (alongside traffic, errors, and saturation).
In depth
Average latency is often misleading because it hides tail latency — the slow outliers experienced by a small percentage of users. The industry standard is to track latency percentiles: P50 (median), P95, and P99. A P99 latency of 2 seconds means 1 in 100 requests takes longer than 2 seconds. High tail latency in distributed systems can be caused by garbage collection pauses, lock contention, slow database queries, network jitter, or cold starts. Distributed traces are the primary tool for diagnosing latency because they show exactly where time is spent across every service in the request path. obseria.io displays latency histograms and percentile breakdowns alongside trace flamegraphs for every service and endpoint.
Related terms
See Latency in action
obseria.io gives you full-stack observability — logs, metrics, traces, and AI-powered root cause analysis.