Incidents & Problems

Separate active symptoms from infrastructure root causes and retained history.

Incidents describe current customer- or service-facing symptoms. Problems describe infrastructure conditions that can explain one or more symptoms. Use both views during diagnosis.

Incidents

The incident table can expose state, severity band, confirmation state, symptom, root, affected hosts, affected customers, revenue at risk, and the highest SLA impact. Open a row for the complete incident detail.

Start with severity and commercial impact, then narrow by symptom, affected objects, and likely root. The Companion helps prioritize; the repair still happens in Uptime.

Problems

Use Infrastructure Problems when several incidents may share a physical or operational cause. A problem can make a pattern clearer than reading each symptom in isolation.

Resolved history

The Companion can retain incidents and problems locally after they disappear from the API. Configure the time window and storage cap in Settings. Retained entries are browser-local and are intended for short-term comparison, not permanent game history.

Correlation is evidence, not an action. Confirm the rack, component, power path, or network path in the game before changing infrastructure.