Settings
Integrations, alert destinations and ingest keys — from dummy data.
Data sources
Where every number on this dashboard comes from
- Dokploy APIProjects, services, deployments and env vars
DOKPLOY_BASE_URL + DOKPLOY_API_KEYlive - Docker socket-proxyContainer metrics and disk usage
DOCKER_PROXY_URLlive - Traefik APIRouter table behind the domain checks
TRAEFIK_API_URLlive - Host metrics (Hostinger API)CPU, disk and memory for the machine itself
HOSTINGER_API_TOKENlive - Uptime history8 monitors reporting
DATABASE_URLlive - Error ingestSentry-compatible envelope endpoint
per-project ingest keylive - Analytics beaconConnected, but low traffic on 1 of 6 sites — worth a look
per-project ingest keydegraded - PageSpeed InsightsNot configured — Quality history stays empty until this is set
PAGESPEED_API_KEYoff
Off means not configured — nothing is broken. Unreachable means configured but silent, and anything depending on it will read as unknown rather than healthy.
Alert destinations
Where a signal goes once a rule decides it's worth telling someone
| Destination | Target | Last test | Status | Actions |
|---|---|---|---|---|
| ops@xyte.be | ok · 2h ago | live | ||
| Pushover | u••••••••••••h9k2 | ok · 2h ago | live | |
| Discord webhook | https://discord.com/api/webhooks/•••/•••• | failed · 1d ago | unreachable | |
| Telegram | not configured | — | off | |
| Generic webhook | https://hooks.xyte.be/panel/••• | ok · 5d ago | live |
Ingest keys
What each client's site or worker authenticates with to push data in
| Project | Key | Created | Last used | Actions |
|---|---|---|---|---|
| Xyte | pk_live_4f1•••••••••••••• | 312d ago | 2m ago | |
| Client B | pk_live_9e3•••••••••••••• | 188d ago | 11m ago | |
| Client C | pk_live_2ad•••••••••••••• | 94d ago | 1h ago | |
| Client D | pk_live_51b•••••••••••••• | 30d ago | 6h ago |
Preferences
Saved in this browser, once wired up
- Refresh intervalHow often every page refetches its source.
- Certificate warningFlag a domain when its certificate expires within this window.
- Disk alert thresholdHost disk turns serious above this. A full disk takes every client site down at once.
- Dense tablesTighter rows, more on screen at once.
- Full hostnamesShow complete domain names instead of compact labels.
- Compact alert wordingShorten symptom labels so incident rows stay scannable.
These aren't persisted yet — reload the page and they reset. Real storage lands once this page is wired to something live.