The check that starts everything else
Every site gets an HTTP or HTTPS check on the schedule you choose — with automatic retries, so a single slow response never gets reported as a false outage.
How it actually checks
A real browser user agent is used for every request, so CDN caching and bot-gating don't hide problems a real visitor would see. The response is read for its status code, timed, and scanned for fatal error strings — a database connection error or a resource-limit message — even when the status code alone looks fine.
- Up to 3 retries on a timeout or connection failure before it's ever logged as down
- Two consecutive failures required before an alert fires — one blip doesn't page anyone
- Response time tracked on every check, flagged separately from a real outage
The details
- Configurable check interval, per site
- Follows redirects, so a domain forward doesn't falsely flag as down
- SSL, port and content checks all run in the same cycle
- Malicious-redirect detection if a site starts pointing somewhere it shouldn't
Questions people ask
Whatever interval you set per site — most customers check every 10-30 minutes.
A non-2xx/3xx status code, a connection timeout, or a known fatal-error string found on an otherwise-200 page.
No — a failure has to repeat on the next check before it's treated as real. A single blip is recorded but doesn't alert anyone.
One location today. If multi-region checking matters for your setup, tell us — it's a fair question and worth a real conversation rather than a marketing answer.
See a real check run.
Bring a URL — we'll show you exactly what a check on it looks like.