How monitoring works

From a provider's announced address space to the figure on its listing.

Monitoring on VPS Advisor is independent network measurement. It is not a provider reporting its own uptime, and it is not a check of one server somebody happened to rent. It is the reachability of a provider's own address space, measured from many places at once by machines the provider does not control.

The chain

1. A provider registers its autonomous systems. An ASN is the identifier of a network on the public internet. This is the one thing the provider tells us, and it is checkable against public routing data.
2. The ASNs are resolved to prefixes. Public BGP data says which blocks of addresses that network announces. This is where the target list comes from: it is derived from what the provider announces to the internet, not from anything it tells us privately.
3. Targets are selected and distributed. A bounded sample of addresses from those prefixes is signed and handed to workers.
4. Workers measure. Independently, from wherever they are. See How a worker operates.
5. Results become a consensus. Many narrow measurements are combined into one verdict per region and one overall. See Worker consensus.
6. The consensus is published. On the provider's listing and its status page, with the network it describes named beside it.

What is being measured, precisely

Whether addresses in a provider's announced space answer, and how quickly. That is a statement about the network path to that provider, aggregated over many paths.

It is not a statement about any particular service. A reachable address does not prove a web server is serving; an unreachable one does not prove a customer's VM is down. What it does capture, reliably and independently, is whether a provider's network is reachable from the rest of the internet, which is the failure mode customers actually notice and the one providers are least able to report on themselves.

Why it is opt-in

A provider chooses whether to be monitored. A listing with monitoring switched off publishes no uptime figures at all, and says so: that is a choice about measurement, not a judgement about reliability, and the page states it in those terms.

Nothing is inferred to fill the gap. A provider that is not monitored does not get an estimated figure, a figure borrowed from a similar provider, or a zero.

Where to read the results

A provider's listing carries a summary card. Its status page carries the full picture: the availability timeline, the regional breakdown, the networks measured, and the incident history. Both name the autonomous system the figures describe, because a reseller and its upstream are different networks behind one brand.