uptimescout
Get an API key

UPTIMESCOUT / Methodology

Methodology

A useful status result explains what was observed, where it applies and when it was last seen.

Fiber and network connectors in a server patch panel.
AI-generated editorial illustration

Four states, with explicit uncertainty

Operational means no issue signal was present in that observation. Degraded means an elevated signal. Outage means a disruption signal. Unknown means there is no usable observation or the state cannot be established. None of these is a personal network diagnosis.

Signals reflect observations rather than a synthetic request from your device. A healthy signal cannot guarantee that every account, feature, provider or city can reach a service.

Regional scope

Observation regions are independent. A result from one region must not be presented as a worldwide verdict. Choose a country in the regional selector to see its current signals and history. Available countries depend on collected observations for each service.

Access restrictions and outage signals describe different things. Regional legal notes do not set the chart state.

Refresh time is not observation time

The interface refreshes once a minute. Collection runs continuously with pacing across services, so a page refresh does not mean that every service has just been observed again. Read observed_at and stale together.

When an observation becomes stale, keep its timestamp visible and treat current availability as uncertain. Do not convert a stale operational observation into an unqualified claim that the service is working.

How grouped history is calculated

The chart uses 36 intervals each for three hours, twelve hours and a day. These are 5, 20 and 40 minute windows. A column shows the most frequent saved signal in its interval. Selecting it reveals per-state observation counts and the last observation time.

An empty interval remains explicitly unknown. We do not invent measurements between observations. Observation counts are sample counts, not numbers of affected users or user complaints. The share of operational samples is not a measured uptime percentage.

What the history can establish

History begins with observations actually saved by UptimeScout. Retention and API access depend on the plan. A signal change can support investigation, but it is not a verified incident start, confirmed recovery time or service-provider incident report.

Content reviewed on 18 November 2026.

Markdown

REST / MCP / SDK

Check a service or integrate the API

Use the service directory for a current regional signal and the documentation for a reproducible integration.

Service directoryAPI documentation