AnchoredIP

Components

Past incidents

Nothing has been recorded here yet.

How this page is produced

It runs on a host that is not part of the service it reports on — a different provider, a different country, and a name served by a different DNS operator — so that it stays up when we do not. It is rendered to static files once a minute; nothing dynamic answers your request.

Most rows are measured from Amsterdam. That is one vantage point, and one vantage point can be wrong about the internet: when a check fails, we first confirm this location can still reach two unrelated networks, and anything we could not verify is published as unknown rather than as an outage. Unknown checks are left out of the percentages instead of counted against them.

The tunnel network, address reputation, provisioning, email and trial availability cannot be seen from outside, so the platform reports those itself. The edge is judged by the reconciliation it records and never by whether it answers a ping — a host that replies while its tunnel interface is down is exactly the failure an external probe would call healthy.

Round-trip latency to our region is a property of the product rather than an incident, so it is not a row here. The measured figures are published on the latency page.

No cookies, no analytics and no third-party requests: this page has to work when everything else does not, and it should not report you to anyone in order to tell you we are broken.