Probe

A probe is an on-demand, one-shot reachability check of a sub-device: the platform asks its gateway to perform a real read on the field bus and report whether the device answered and the round-trip time. The platform only relays the request; it never interprets or stores the reading as business data.

Probe vs presence

Presence is the platform's continuous, cached connection state. For a passive sub-device it reflects only what its gateway asserts while the gateway's own platform session is alive — a device that drops off the field bus (power loss, wiring fault) can still appear online until the gateway notices via a health check. A probe instead proves end-to-end application reachability at the moment it runs and returns the measured round-trip time. It never updates presence or any other device state.

Probe vs health check

• Probe — manual, one-off batch check; results are shown only to the operator.
• Health check — a periodic probe loop configured on a connector instance that automatically drives the bound device's online/offline state.

Probe modes

The platform only validates and forwards mode; its precise semantics are defined by the edge implementation.

Mode	Recommended meaning
cached	Answer from the most recent known check; no new field-bus read is required.
fresh (default)	Perform a new read on the field bus now.
strict	Perform a new read under the strictest checking rules.

A probe targets only sub-devices bound to a gateway. Devices on an offline gateway are not queued; they fail immediately with a not-connected error.