Skip to content

Monitoring

UNITRONIX edited this page Sep 6, 2026 · 1 revision

Monitoring

How to watch your BetterDesk server: Prometheus metrics, health checks, and audit logs. Data stays on your host — see Privacy.


Health check

curl -s http://localhost:21114/api/health
curl -s http://localhost:21114/api/server/stats

Expect JSON OK / peer counts. stats uses the same access rules as /metrics (auth or allowlist).


Prometheus metrics

Go serves Prometheus text on the admin API port:

http://localhost:21114/metrics

Access is gated (API key / auth, or IP allowlist / public flag via METRICS_* env — see server config). Do not expose /metrics to the open internet without an allowlist.

Point Prometheus at that URL; Grafana uses a normal Prometheus datasource.


Audit log

Auth and device actions land in your audit store. Review in the panel (Audit) or:

curl http://localhost:21114/api/audit \
  -H "X-API-Key: your-api-key"

Panel device metrics

CPU/RAM/disk in the console come from clients reporting to your API. That is fleet ops, not vendor analytics. Quiet it down with enrollment sync modes — Privacy.


Shared NAT / CGNAT

Many peers behind one public IP: see Settings for ALLOW_SHARED_NAT_INITIATOR (default off) and set MASK / -mask to your real LAN CIDR (empty keeps historic /24). Details in 3.5.x release notes and Troubleshooting.


See also

Clone this wiki locally