Monitoring, on your terms
SiteSonar exists for a simple reason: the people running the infrastructure should own the data about it. It is self-hosted monitoring for servers, websites, uptime checks and AI agents — watchful by default, and quiet until something actually needs you.
Born from open source, rebuilt for the whole stack
SiteSonar grew out of the open-source nMon monitoring app. We rebuilt and rebranded it, then kept going — adding deeper Linux server monitoring, per-site web-traffic analytics, and a brand-new, generalized agent platform with a light/dark Fleet Console.
Everything flows through the unified agent layer — push a number from any server, service or script and watch it, graph it and alert on it the same way, all from one console.
- Vanilla PHP. No framework, no build step, no node_modules — just PHP and a database you already know how to run.
- Runs on your own LAMP stack. Dependencies are vendored, so it drops onto the box you already have and starts watching.
- No telemetry. Nothing phones home. Your hosts, your metrics and your alerts stay entirely on your infrastructure.
- Roles, permissions & groups. RBAC and groups scope exactly what each person sees — handy for teams and clients alike.
A few opinions, baked in
Every design decision in SiteSonar comes back to three ideas — and they are the same ideas whether you are watching one server or a hundred.
Self-hosted by default
Your monitoring data never leaves your own servers. No cloud account, no third party in the loop — you own the box and you own the data.
Simple over heavy
No build step, no framework churn, dependencies vendored in. Boring, legible code you can read, deploy and trust for years.
Catch it before users do
Set tolerances once and let confirm-after-N debounce filter the noise — so the alert that does reach you is the one worth waking up for.