Monitoring
API monitoring
HTTP checks on your internal services and endpoints, from inside your LAN.
The probe polls HTTP/HTTPS endpoints — REST, GraphQL, health endpoints, login pages — from inside your network, so you're testing what your users actually reach, not what the public internet sees.
Assertions
A request failure marks the check down; a failed assertion marks it failing. Both alert — at different severities.
| Assertion | Fires when |
|---|---|
| Status code | the response code differs from what you expect |
| JSON path | a value in the response body doesn't match (e.g. $.status == "ok") |
| Latency | response time exceeds your threshold |
Noise control
- Alerts fire after 3 consecutive failures — one timeout doesn't page anyone
- Recovery alerts fire on the first success, with the outage duration
- Optional device link: respects that device's alert window and group routing