Understand monitoring rules

Healthy means the configured checks passed

A healthy result means the latest check passed the monitor's configured expectations. It is not a guarantee that the website or email service is working or secure. A monitor without a completed check is not yet confirmed healthy.

Presence rules require a record of the selected type. Expected-value rules check intended answers. Mail-specific rules inspect policies such as SPF. Choose a rule that describes what matters for this resource.

Answers can legitimately differ

The default presence-monitor behavior requires the authoritative nameservers to agree while allowing public recursive and DNS-over-HTTPS answers to differ. Cached answers, geographic routing and CDNs can cause differences; an observed difference alone does not identify its cause.

For strict comparison, choose All DNS sources must agree. For intentional variation at authoritative servers too, choose Answers may vary across all sources. Missing records, failed queries and explicit expected values remain errors. Explicit agreement rules still require agreement in their selected group.

1. Example: after an intended IP change, a public resolver may briefly return an older cached address.
2. Compare the source answers and timestamps. Confirm that the authoritative servers have the intended values.
3. Allow variation only when it fits your service. Do not use it to hide missing records or repeated DNS failures.

Check interval and severity

The interval controls how often checks are scheduled. Incident severity describes the importance of a failure, not the current health: a healthy monitor can have Incident severity: warning.

Use confirmation thresholds to avoid alerting on a single temporary failure. Changing rules, sources, thresholds or enabled state resets confirmation counters.