1 · Poll
Fetch every device & monitor across all consoles, every 2 min.
2 · Confirm
Flap protection, must stay down 2 cycles before filing.
3 · File
One incident → one ticket, categorized & tagged.
4 · Recover
Clears on its own → ticket auto-closed with a note.
Live device feed
Knock a whole property offline and watch the engine confirm it across poll cycles, then file a single property-outage ticket, not one per device. Bring it back and the ticket auto-recovers.
Asana ticket queue 0 open
No open tickets, all systems nominal.
Faithful demo of the Engine service. The real pipeline runs as a scheduled function every 2 minutes, is idempotent (each ticket carries a tag so it can never be duplicated), and is fail-safe (a total source outage writes nothing rather than mass-closing tickets). Time is compressed here so a "cycle" is ~2 seconds.