Fajita requests your pages the way a visitor would, on a schedule, and treats every slow or wrong answer as a lead worth chasing.
What it checks
Availability, HTTP status codes, and response time for any public URL. Each check is a real request, not a ping.
What failure looks like
A timeout, a connection refused, a 500 where a 200 belongs, or a page that answers in four seconds instead of two hundred milliseconds.
What Fajita does
It re-checks before saying a word. A confirmed failure opens an incident, alerts your channels, and starts the clock on your uptime record.
Why it matters
Most small teams learn about downtime from a support ticket. The ticket arrives after the damage. A monitor arrives before it.
In practice
Fair questions
No. A failed check triggers verification, not an alarm. You hear about confirmed problems only.
No agent, no SDK, no code change. Fajita checks from the outside, the same direction your customers arrive from.
Fajita verifies the failure, opens an incident, alerts your configured channels, and sends one clear recovery message when checks pass again.
Related