Public network
Network status
Every server in Chirp's public network, probed from outside every 30 seconds. No figure here is something a server says about itself.
Loading the figures…
No servers yet
The public network starts with its first node. Until then, Chirp works with a relay you run yourself, or one someone you trust runs.
How these figures are made
- Probing
- The monitor asks each server for /healthz over HTTPS every 30 seconds, from a host outside every server it watches, on a fresh connection each time. A server reports nothing about itself, and the relay's health check reveals no state.
- Uptime
- The share of the minutes a server was probed in that it was up. A minute counts as up when most of its probes succeeded; a tie counts as down. Figures are cut, never rounded, so 100% means every probed minute. The 30- and 90-day figures count whole UTC days, today so far included.
- No data
- A minute nobody probed, because the monitor itself was down, counts neither as up nor as down. In the 90-day strip it is a blank day.
- Off-line
- Its last three probes failed: about 90 seconds without an answer. The time shown runs from its last answer.
- Failing
- Its latest probe failed, but not yet three in a row.
- Unknown
- The monitor has not probed it for three rounds, so this page does not say what it is doing.
- Dedicated
- A Chirp relay in public mode that has joined the network, run by the operator named on it.
- New
- Joined the network less than 30 days ago. Its 90-day figure is still a short history.
- Suspended
- Still listed, but broke a rule of the network, and left out of the list Chirp uses until it is fixed.
- Latency
- How long the latest probe took, when it succeeded, including a fresh TLS handshake. It depends on where the monitor is.
The rules, and how a server joins the network, are in the proposal. The raw figures are in the API.