Glossary
Alert
A signal that a condition was met and someone should look — the input to incident response, not the response itself.
An alert is a signal raised when a monitored condition is met: a check failed, a threshold was crossed, a heartbeat did not arrive. It is an observation, not a conclusion. Whether it means something is broken, whether anyone should be woken, and whether it belongs with other alerts are separate decisions that come after.
Three words get used interchangeably and mean different things. The ALERT is the raised condition. The INCIDENT is the tracked disruption an alert may open or join. The NOTIFICATION is the delivery — the email, the Slack message, the phone call — that carries either to a person. Keeping them apart is what lets one incident collect twenty alerts and still result in one phone call.
The useful test for whether a condition deserves to be an alert is whether there is something a person would do about it, now, that they would not otherwise do. Conditions that fail that test are still worth recording and worth charting — they just belong on a dashboard, where they inform, rather than in an alerting pipeline, where they interrupt.
In Uptimely
Uptimely treats alerts and incidents as separate records with parallel lifecycles: a monitor's criteria can raise an alert without declaring an incident, both auto-resolve when the condition clears, and both are subject to the same mute, dependency and failure-threshold rules.
Part of the Uptimely monitoring glossary — plain definitions of the vocabulary uptime monitoring, incident response and status pages are discussed in. Something here read wrong to you? Email support@getuptimely.com.