Six regions globally: us-east, us-west, eu-west, eu-central, ap-southeast, ap-northeast. All on Cloudflare's edge network. See Regions.

A monitor checked every 60 seconds typically alerts within 90 seconds of going down — one missed check + multi-region confirmation + dispatch latency. For 30s checks, alerting is typically under 60 seconds.

A single region failing usually means a network blip or a CDN edge issue, not your service being down. Requiring 2+ regions to fail before alerting eliminates ~95% of false positives.

Not directly with HTTP/TCP monitors — they run from the public internet. Use heartbeat monitors: have your internal job ping /api/heartbeat/<slug> on a schedule. If we don't hear from it within the grace period, we alert. See Heartbeat guide.

Not as a packaged release. Enterprise customers can run a private edge instance. Talk to us.

See happyuptime.com/status/happyuptime for our public status page. We use a separate provider (different cloud, different region) for our status page so it stays up even if our primary stack is down.

Yes. Monitor → Pause stops checks and suppresses alerts. You can also create a planned-maintenance incident on a status page that suppresses alerts for the linked components for the maintenance window.

On-call alert channels resolve the current on-call person at page time (override → fixed shift → rotation, gated by restrictions), then @mention them in the channel and DM them. See Slack paging.

Yes — that's what layers are for. Primary + Secondary + Manager is a common pattern. All paged in parallel, or in escalation order. See Multi-layer schedules.

Yes. Each schedule has a personal iCal feed. See Calendar integration.

A monitor detects a problem. An incident is the human-managed record of that problem (with timeline updates, assigned responder, severity). A status page is the public-facing presentation. One incident can affect multiple status page components.

Yes — checks make real HTTP requests. At a 60s interval across 3 regions, that's ~4,300 requests/day per monitor. For high-traffic origins it's noise; for budget-sensitive endpoints, increase the interval or use heartbeat monitors instead.

Yes. happy config pull > backup.yml exports all monitors, status pages, alert channels, and on-call schedules. Check history is exportable from Settings → Export → Request CSV (delivered by email).

All data lives in Cloudflare D1 (SQLite at the edge), replicated globally. EU customers can request EU-only storage on Enterprise plans.