How does alert deduplication work?

Beacon groups related failures into a single incident instead of paging you once per check.

• Consecutive failures of the same monitor extend the open incident — they never create a new one
• Failures across different monitors within 120 seconds that share a failure region are grouped under one incident with multiple affected components
• Recovery requires 3 consecutive successful checks, which prevents flapping monitors from re-paging every minute

You can tune the grouping window (30s–10min) per escalation policy on Team plans and above.