Alert Policies

Overview

Alert policies control when and how MonitorHound sends notifications. They determine how failures are confirmed and when alerts are triggered.

Failure confirmation

To reduce false positives, MonitorHound confirms failures using multi-location checking:

1. Each check runs simultaneously from all selected locations
2. The evaluator aggregates results across all locations
3. If the monitor transitions to an unhealthy state, alerts are sent to all assigned channels

This multi-location checking eliminates false alerts caused by temporary network issues between a single monitoring location and your server.

Alert timing

Alerts are triggered immediately when the monitor’s state changes from healthy to unhealthy. There is no delay — as soon as the evaluator determines the monitor is unhealthy based on check results from all locations, the alert fires.

Recovery notifications

When a monitor that was in an unhealthy state returns to healthy, a recovery notification is sent to the same channels that received the failure alert. Recovery notifications include:

• Monitor name and type
• Previous state and new state
• Check results from each location

Notification frequency

During an ongoing outage, MonitorHound does not re-send alerts for every failed check. Instead:

• One failure alert is sent when the outage begins
• One recovery alert is sent when the outage ends

Best practices

• Assign different channels for different severity levels — email for non-urgent, SMS for critical (when available)
• Keep alert channels up to date — remove channels for team members who have left
• Test your channels regularly using the Test Alert button on the monitor detail page to verify delivery