mirror of https://github.com/UNITRONIX/BetterDesk.git synced 2026-09-10 17:45:42 +00:00

Files

T

UNITRONIX 868aa5b22f docs: slim README, refresh wiki for 3.5.x, add Privacy

Replace the monolithic README with a short landing page, align wiki
with current 3.5.x product facts, add PRIVACY.md, Monitoring and
Unattended/WoL guides, and park historical audits under docs/archive.

Thanks: INSOLVE (Honorary); Marco Jakobs (@jacotec); MyNameisStitch (@MyNameisStitch); Redspin (@playerumpknow)

2026-09-06 12:25:34 +02:00

1.5 KiB

Raw Permalink Blame History

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.

1.5 KiB Raw Permalink Blame History