Locations

Where a check runs from.

A location is a named place checks run from — for example local, or a geography you add with a remote worker. Each location is served by one or more workers, and every result a worker reports is labelled with the location it came from.

Manage locations in the dashboard. Assign a monitor to specific locations from the dashboard, or over the REST API and Python SDK.

Called regions in the API. The REST API, the Python SDK and workspace import/export still use the field name regions, and it is not changing — renaming it would break every existing client. The dashboard says locations; the wire format says regions. They mean the same thing.

A monitor with no location is never checked — it stays at “No Data” forever. Assign at least one location to every monitor you want watched. This applies to monitors created in the UI, over the REST API, and via import.

Why more than one

A single location tells you whether it can reach your site. Several tell you whether the site is actually down, or whether one vantage point is having a bad day — which is the difference between a real incident and a false alarm. How many have to agree is a per-monitor choice; see How many locations must fail.