System status
Live view of the infrastructure that watches your websites.
About this status page
This page shows the live operational state of the infrastructure that watches your websites: the independent monitoring regions that run uptime and security checks around the clock. Each region reports in continuously, and the view here refreshes every minute.
When your own site looks down, WWT re-checks it from more than one region before alerting you, so a network problem near a single vantage point never becomes a false alarm. If a region itself goes offline it appears here as Down, and its verdicts are set aside until it recovers.
How a check works
- Every monitored website is requested from an independent monitoring region on its own schedule, the way a visitor's browser would request it. The region records whether the site answered, how long it took and what came back.
- A failed check is never final on its own. Before anything is reported, other regions repeat the same check, and the failure is only accepted when they agree with each other.
- Only a confirmed failure opens an incident and sends alerts. A recovery is reported the same way, once the site answers again from the regions that saw it fail.
The exact definitions - what counts as a failure, when an incident opens and closes, how availability is calculated - are on a page of their own: How we measure availability
Why we raise fewer false alarms
Most monitoring tools alert the moment a single probe fails once. A congested link near that probe, a brief DNS hiccup or a routing change between two providers then becomes a page at three in the morning about a website that was never down. We treat one failed check as a question, not as an answer.
Three mechanisms work together. A failure has to repeat a configurable number of times before it counts. It has to be confirmed from more than one region, so a problem local to a single vantage point is filtered out. And a region whose results are wildly out of line with the others for the same round is treated as the faulty party, and its verdicts are set aside for that round.
Regions check themselves
A monitoring region that has lost its own internet connectivity would otherwise report every website as down. To prevent that, each region runs a canary self-check: every 30 seconds it reaches out to a set of diverse, highly available anchors on the public internet. If those anchors become unreachable, the region declares itself degraded.
A degraded region keeps running, but its verdicts are excluded from every decision until the canary passes again. Independently of that, every region reports in to WWT continuously; a region that goes silent is marked unavailable and shown as Down on this page. Either way, a broken vantage point can never be mistaken for a broken website.