★ this host (appliance, shuts down last) · mains on battery shutdown triggered unreachable

Test shutdown: simulates the shutdown order without effect. Reset state: clears a test/alarm latch.

Monitor one or more UPS devices — via SNMP (UPS with its own network card) or via a NUT server, which is how USB and serial UPS devices are read. For servers with redundant power supplies use one UPS per feed — the mapping is done in step 2. Optional per-UPS thresholds override the global values from step 3.

API token with the Sys.PowerMgmt privilege. The host marked ★ “This host” (the appliance) is shut down last.

One matching condition is enough. Empty field = condition disabled.

Empty by default (fail-safe): an SNMP loss triggers no shutdown, because it is usually a network/UPS defect and not a power outage. Only set this when a prolonged communication loss should deliberately be treated like an outage — the actual power state is unknown then.

Time and timezone are controlled by the Docker host / orchestrator, not by this container.

Save the configuration as a JSON file or restore it — ideal for reinstalls or migration.

The export file contains credentials in plain text (SNMP community, API token). Store it safely!

Current version: –. Upload a release package (.tar.gz/.zip); the system applies it and restarts the service.