Every dot below is a real, continuously repeated network check -- not a simulation and not a static promise. Where we have an active customer VPS in a region, we check that instance directly; everywhere else, we run a live network-health check for that metro so every region still shows something real. We don't name the infrastructure behind any region here or anywhere else -- see our Refund and Service Level Policy for why we don't offer a fixed contractual uptime percentage.
A live feed of the checks above as they actually happen, most recent first.
Every region on this page is checked on a rolling cycle using real internet measurements from a third-party network-probe service, not from a machine we control -- so the result reflects genuine, independent reachability, not our own self-reporting. A region shows Reachable once its most recent check succeeded, Check failed if we have an active customer instance there and it didn't respond, and Awaiting first result if it hasn't been checked yet. Results refresh continuously in the background.
Regions without an active customer instance yet are checked against a neutral, independent internet target instead, since there's no server of ours there to probe directly -- a useful general signal, but not a perfect one. Some networks (national firewalls being the most common cause) block that neutral target for reasons that have nothing to do with our own infrastructure. Rather than mislabel that as a real outage, we show it as no confirmed signal instead of "down," and it's excluded from the uptime percentage above.