Monitoring
HTTP Monitoring
Probe a URL on a schedule and assert status (and timing) from the real response. Available on Free and above.
Overview
An HTTP monitor (kind http) issues a GET or HEAD request to a public HTTPS/HTTP URL you configure. Pass/fail is derived from the probe outcome — expected status, timeouts, and network errors — not from synthetic uptime.
Configuration
url— target URL (SSRF-safe hostnames only; private / link-local targets are rejected).method—GET(default) orHEAD.expectedStatus— HTTP status code to treat as success (default200).timeoutMs— probe budget (default10000, max 60s).followRedirects— defaulttrue.
Create and edit monitors under the project Monitoring UI, or via the REST API when Enterprise api_access is enabled.
Results
Each scheduled run stores a check result. Failures (wrong status, timeout, network error) contribute to health and can open alerts. Continuous runs require Trigger workers — see Monitoring overview.
Plans
HTTP monitoring is available on Free and all paid plans (feature gate http_monitor). Project and schedule limits still apply per plan.
Limitations
- Only GET and HEAD — no request bodies or custom auth headers.
- Unsafe or non-public hostnames are blocked (SSRF protection).
- Empty check history means no successful scheduler run yet — not “100% uptime.”