METHOD
How we score status
One long-running regional incident shouldn't make a whole provider look down. Here is exactly how the headline is decided.
How we score status
- Source. Every reading comes from the provider's own public status page or status API (Statuspage, Instatus, incident.io-style JSON, or the provider's RSS/JSON). ORBIT never pings or scans the service.
- Scope. For each open incident we count the regions or components it names against the total the provider publishes, and label it global, multi-region, 1 region or n of N components.
- Age. Incidents are new (under 1 hour), active, long-running (open more than 48 hours) or stale (no provider update for 24 hours or more).
- Headline. The headline colour comes from current incidents only. A single-region or small-scope incident caps the headline at "Degraded"; a global or widespread one keeps the provider's impact. Long-running or stale issues limited to one region or a few components get a known long-running issue badge and don't set the headline colour.
- Transparency. The provider's own page-level status is always shown next to ORBIT's headline, with a link to the official page. When they differ, the provider's page is the authority on what it says.
Current thresholds
- New: open less than 60 minutes
- Long-running: open more than 48 hours
- Stale: no provider update for 24 hours
- Limited scope: one region, one component, or less than 25% of the regions/components the provider lists