Incidents
What an incident is in ARGUNIS, how it opens, what statuses it has, and what you can do with it.
On this page
An incident is the ticket ARGUNIS opens when something you monitor fails. The platform correlates events: an alarm opens the incident, repeats are added to it, and the recovery closes it. That way you don't get a notification for each event, but one per problem.
How an incident opens
You don't need to open them manually. ARGUNIS opens an incident when an alarm arrives from any source: a web or TCP service down, a server not responding, CPU, RAM, or disk above the threshold, a failed UX test, a database, a cloud service, an APM error, an agent or probe finding, or a security detection, among others.
Before opening it, ARGUNIS checks the Alarm policies: a policy can require the alarm to persist for a few minutes, to exceed a threshold, or to stay as an event only. Without policies, every alarm opens an incident.
Each incident gets a ticket number (for example, INC-00012), a severity (Critical or Warning), and a Resolver groups according to the Notification rules and channels.
A host marked as "in maintenance" or "inactive" in its settings doesn't open incidents: its failures are kept as events only.
Statuses
Status | What it means |
|---|---|
Open | Nobody has acknowledged it yet. It escalates if nobody takes it |
Acknowledged | Someone is handling it. It no longer escalates |
Resolved | The problem ended: it recovered on its own, closed due to silence, or someone closed it |
An unresolved incident can also be snoozed until a specific time: meanwhile, it is not re-notified and doesn't escalate.