Using the dashboard
The dashboard is designed so anyone — not just engineers — can answer one question fast: is everything OK, and if not, what do I fix? Open the live demo to follow along.
The overview
The home page shows every server as a card, color-coded by status, with a count of each status across the top.
- OK — nothing needs attention.
- WARNING — something is trending the wrong way (e.g. a disk past 80%).
- CRITICAL — act now (e.g. a disk past 90%, about to fill).
- STALE — the agent hasn’t reported recently; the server may be down or offline.
Each card shows memory and top-disk usage, how many services are tracked, and — if there’s a problem — the reason in plain English, like “disk /data is 94% full.”
How status is decided
| Signal | Warning | Critical |
|---|---|---|
| Disk usage | ≥ 80% | ≥ 90% |
| Memory usage | ≥ 90% | — |
| No recent report | — | Stale after 5 min |
The worst signal wins: one critical disk makes the whole server critical. Custom thresholds and checks are on the roadmap and via plugins.
Server detail
Click any server to see the full picture:
- Overview — CPU, cores, uptime, load average, package count, last report time.
- Memory — used vs. total with a usage bar.
- Disks — every mounted filesystem with a usage bar.
- Services — what’s running vs. stopped, in a searchable list.
The API
Everything on the dashboard is available as JSON for scripts and integrations:
curl https://monitor.example.com/api/v1/servers