Telemetry
Optional anonymous usage counts, off by default.
A self-hosted Dock sends nothing about itself unless you turn telemetry on. It is off by default. When it is on, the control plane sends one small report about a minute after boot and then once every 24 hours.
Turn it on
Set this in the backend environment and restart:
DOCK_TELEMETRY=onSet DOCK_TELEMETRY=off, leave it empty, or remove the line, to stop. DOCK_TELEMETRY_URL
overrides the endpoint (default https://api.usedock.io/v1/telemetry).
What is sent
The report is a JSON POST with exactly these fields and no others. The
counts are totals for the whole instance, across every organization and
workspace:
| Field | Type | Meaning |
|---|---|---|
version | string | Control plane version |
os | string | Operating system of the control plane (linux, darwin, win32) |
arch | string | CPU architecture of the control plane (x64, arm64) |
hosts | number | Number of registered hosts, online or offline |
docks | number | Number of docks that exist, in any state (archived included; deleted docks are not counted) |
runningDocks | number | Number of docks in the running state |
Example:
{
"version": "0.1.0",
"os": "linux",
"arch": "x64",
"hosts": 2,
"docks": 14,
"runningDocks": 3
}What is not sent
There is no instance id, and no organization, user, dock, host or workspace names. No IP addresses and no URLs are included in the report body; like any HTTPS request, the endpoint sees the address the request comes from. Requests have a short timeout; a failure is ignored and never affects the control plane.