One incident, not thirteen alerts
The uplink on dist-sw-03 went down at 03:41. Twelve access ports and two access points went with it. DeviceWatch groups them: one incident, the cause named, the symptoms listed underneath. Thirteen separate emails at 3am is how alerting gets muted.
It learns what normal looks like
Thresholds you set by hand are wrong twice — too tight on the busy link, too loose on the quiet one. DeviceWatch builds a baseline per port and per device and alerts on the departure from it, so a 40% link that is normally 4% is worth telling you about.
Plain SNMP, nothing installed
Read-only SNMP v2c or v3 against the devices you already have. No agent, no collector VM, no port mirror. If it answers ifTable, DeviceWatch can watch it.
Thirty days, then a key
The trial is the whole product with no device cap — discovery, the topology map, forecasting, webhook alerts and the scheduled PDF report included. DeviceWatch has no free tier: when the trial ends, polling stops and the window asks for a licence. Nothing you collected is deleted, and entering a key brings the history straight back.