Every running container is checked twice a day — the gateway responds, the agent is alive, TLS is valid and the disk is not full.
Once a day we create a real station from scratch — provision, DNS, TLS, agent install, health check — confirm it works, then tear it down. If signup is broken, this finds out before you do.
Measured per station by an independent uptime monitor that requests your station’s health endpoint from outside our network. Planned maintenance and time a station spends stopped at your own request are excluded; everything else counts against us.
This page reports shares, never counts. We don’t publish how many stations we run, and we never name one — not the hostname, not the owner, not what it’s connected to.
If the health check itself stops reporting, this page says the status is unavailable rather than continuing to show yesterday’s green. Something specific to your own station that isn’t reflected here? Tell us — a fleet that is 100% healthy on average can still be broken for you.