Uptime for every device the token and the device filters select, over one
window [start, end), computed from every heartbeat_status_change event
in the window. Per device: uptime (online seconds / window seconds; At
Risk counts against it), seconds and shares per status, transitions, and
the status at the window start and end. Also a summary (devices in scope,
never reported, evaluated, matching, an uptime histogram and the mean) and
paging.
A window with no status change is filled with the status known at its
start: the heartbeat_last_status of the first event after the window,
else the device's current status. Any part of a window before the device
partition's device_events retention cutoff is no_data (a device whose
current status is never_reported stays never_reported).
Devices that had not reported by the window end are excluded from the
evaluation and counted in summary.never_reported.
Device filters are the ones GET /devices accepts, as
<field>_<operator>=<value> (for example device_type_id_eq,
tags_incany, heartbeat_status_in); the most used are listed below.
heartbeat_status_* filters use the raw devices-table vocabulary
(online, offline, idle, never_reported). A stats key under an
operator this route does not support (for example uptime_eq or
start_gte) is a 400, not an ignored filter.
| Time | Status | User Agent | |
|---|---|---|---|
Retrieving recent requests… | |||
Invalid window, order_by, paging, bucket, status or result-filter value, a stats key under an unsupported operator, or unparseable device filters.
You are not authorized
An unexpected error was encountered.