Help in under five minutes

Monitor, understand, act.

This documentation follows the product itself. Start with a check, create your monitor, then open its details when an anomaly appears.

Check a website Read the guide

01

Create your first monitor

  1. Check the domain. The public result confirms that the URL is valid and reachable from our probe.
  2. Open your member area. Add the full URL, give it a short name, and choose the interval.
  3. Wait for the first check. The details refresh automatically; do not recreate the monitor while its state is still unknown.
Mobile tipUse a complete HTTPS URL. The keyboard and create button remain available without horizontal scrolling.

02

Understand states

Operational

The probe received a response considered healthy.

Degraded

The service responds, but its status or latency needs attention.

Unavailable

The connection failed or the response confirms an error.

Unknown

No usable measurement is available yet. This does not mean the service is up.

Paused describes the check lifecycle, not website health. A paused monitor no longer produces measurements.

03

Read history

  • The availability timeline keeps the most severe state when several checks are grouped.
  • The average describes the general level; P95 reveals the slowest 5% of checks.
  • Change periods without losing context: the selected period remains shareable in the address bar.
  • An empty period must remain unknown and must never be counted as available.

04

Start with a rule template

The Essential HTTP, SEO, commerce, blog, and security templates add editable deterministic rules. Their version number identifies the exact definition used as a starting point.

  1. Choose a template from the monitor Configuration tab.
  2. Review the added rules: they remain a draft and are not saved automatically.
  3. Adjust thresholds and disable checks that do not fit the page, then use Test now.
Keep in mindA template does not prove that a site is correct. It provides an explainable baseline without JavaScript or automated AI decisions.

05

Interpret regional probes

Oracle Paris, Google, and Globalping results are independent observations. A failure in one region can indicate routing, DNS, filtering, or CDN issues rather than a global outage.

CompareExact time, HTTP status, latency, error details, and whether at least one other region is operational.

06

Resolve “no measurement received”

  1. Check that the monitor is active and its next run is scheduled.
  2. Choose “Run now” and wait a few seconds before refreshing.
  3. If the main result exists but regions do not, check the Oracle and Google services and their MongoDB access.
  4. Compare UTC timestamps before concluding that data was lost.
Check service health