Live dashboard of resources per site, high-resolution MRTG history (1 min, up to 1 year), and TTFB test from your server, an external SSH client or geolocated worldwide nodes (EU, US, Asia). You measure what your customers see.
A single docker stats call, 5 to 60 s cache, concurrency lock: only one collector runs even if 20 admins watch the dashboard. Minimal server load.
Per-container stats: CPU%, memory used, network I/O, disk reads/writes. Shown in the dashboard and each domain card.
Direct measurement of "accepted conn" via cgi-fcgi on pm.status_path. Captures every PHP request, even those < 200 ms invisible to snapshots.
Minute-level cron collector. Views: 1h, 24h, 7d, 30d, 1 year. Resilient: skips if the previous run is still going, 25 s timeout on docker stats, traffic snapshots never lost.
3D visualization of containers, 5 s refresh, sphere per site, height = traffic, color = health. The rest of the dashboard is cached 60 s to keep load minimal.
Configurable email alerts on CPU, RAM, disk, restart count. Anti-spam rate limiter: one email per threshold per cooldown window.
Per-domain HTTP probe, configurable path, history with downtime windows. Fast incident detection.
3 test modes: local (server), SSH (from an external client), external multi-region via check-host.net. DNS, TCP, TLS, starttransfer, total phases — captured by curl -w.
Curl runs on the host, measuring the internal stack. Detects PHP / DB / cache regressions. 1 to 10 iterations, min / median / max / mean aggregates.
Wrapped in ssh user@host: curl runs OVER THERE. DNS / TCP / TLS phases become realistic — you see what a visitor would go through.
Geolocated probes via check-host.net: Europe, Americas, Asia or worldwide. 5 to 10 nodes per region. No auth, no token.
We run a TTFB test on one of your domains from 3 continents during the demo. No commitment.
