Browse docs

What is uptime monitoring

How WatchCat checks your endpoints and decides when something is actually down.

Uptime monitoring means regularly checking that a URL or API endpoint is reachable and returning a valid response. Instead of waiting for a customer to report an outage, you get an alert within minutes of a failure.

WatchCat sends HTTP requests to your endpoints on a schedule you choose. Each check records the status code, response time, and response body. When checks start failing, WatchCat opens an incident and notifies your team.

How checks work

At each interval, WatchCat sends a GET or HEAD request to your monitor URL and evaluates the response. By default, a check is considered successful when the response status code is in the 2xx range. You can configure which status codes count as healthy — single codes, ranges like 200-204, wildcard families like 2xx, or comma-separated combinations.

You can configure the check interval: every 1, 2, 3, 5, 10, or 30 minutes. Shorter intervals mean faster detection; longer intervals are appropriate for endpoints where occasional latency is expected.

Failure detection and confirmation

A single failed check does not immediately open an incident. Networks are noisy — a request can time out, a server can briefly restart, a CDN can return a transient error. Alerting on every individual failure would produce too many false positives.

Instead, WatchCat uses a configurable confirmation period. Failures must continue for this entire window before an incident opens. During that time, failures are recorded but no alert is sent. If the service recovers on its own, the monitor returns to OK with no noise.

Example: With a 5-minute confirmation period and a 1-minute check interval, a monitor won't open an incident until checks have been failing for at least 5 minutes — filtering transient blips while still alerting quickly on real outages.

Incidents

When a monitor transitions to DOWN, WatchCat opens an incident. The incident captures the start time of the outage and triggers your configured alerts.

WatchCat continues checking during an active incident. When checks start passing again, the monitor transitions to OK and the incident is resolved — recording the total downtime duration.

Every incident has a full timeline: when it started, when each check ran during the outage, which alerts were sent, and when the service recovered. You can review past incidents from the dashboard at any time.

What to monitor

Good candidates for uptime monitoring:

  • Your main application URL or login page
  • Public API endpoints that customers depend on
  • Webhook receivers or payment callbacks
  • Internal services your team accesses
  • Health check endpoints your infrastructure exposes

Start monitoring in minutes

Free plan available. No credit card required.