DNS Diff monitors
Detect any unplanned change to a DNS answer — hijacks, mistaken edits, and provider drift.
A DNS Diff monitor records the answer it gets, then compares every later answer
against it. When the answer changes, the monitor moves to the CHANGED status
and alerts fire.
It is a security and change-detection control, not an availability check.
Configuration
| Setting | Allowed values | Default |
|---|---|---|
host |
the name to watch | required |
nameServer |
a specific resolver to ask | system resolver |
There is no expected value — that is the point. The baseline is whatever the first successful lookup returned.
The CHANGED status
CHANGED is deliberately sticky. It does not clear itself when the answer
settles, because “the record changed and then stopped changing” is exactly what
a successful hijack looks like. A human has to look at the new value and accept
it:
POST /monitors/<monitor-id>/dns/acknowledge
Acknowledging adopts the current answer as the new baseline and returns the
monitor to UP. In the dashboard this is the Accept change action on the
monitor.
What it catches
- A registrar or DNS account compromise repointing your domain.
- A colleague editing a record without telling anyone.
- A provider silently changing the addresses behind a managed record.
- A failover that fired when you did not expect one.
Expect legitimate noise
Some records change constantly by design and are a poor fit:
- Records behind a CDN or global load balancer that rotate addresses.
- Very low-TTL failover records.
ANYlookups against providers that reorder answers between queries.
Watch the specific record type you care about rather than the whole name, and
prefer names whose values are meant to be stable — your apex, your NS, your
MX.