Health Checks & Status

A server can have 10% CPU usage and still be completely broken if the application daemon has crashed or the database is refusing connections. Raw resource metrics alone are not enough to know that your service is actually functioning.

RasnuOS solves this through native, customizable Health Checks that monitor your application at the layer that matters: the one your users actually hit.

Configuring Health Checks

Health Checks are configured per-server and per-Server Group directly from the dashboard. You do not need to install any third-party monitoring agent.

The following check types are supported:

• HTTP / HTTPS: The Control Plane sends an HTTP request to a URL you specify and validates that the response status code falls within an acceptable range (e.g., 200–399) within a defined latency threshold (e.g., under 500ms). This is the most common check type for web applications.
• TCP Port: Verifies that a specific TCP port on your server is actively accepting connections. This is ideal for monitoring databases, message queues, or any service that doesn't speak HTTP.

How the Check Works

Health Checks are executed from RasnuOS's external monitoring nodes — not from within the server itself. This is a critical distinction. An internally-executed check could pass even if the server's public network interface is completely unreachable.

When a configured check fails:

1. The failure is immediately recorded as a permanent, timestamped Health Check Result, building an auditable history of your service uptime.
2. The status indicator on the Health Dashboard updates in real-time.
3. If you have configured Alert channels (Slack, Email, PagerDuty), the alerting engine fires instantly.
4. If you have AI AutoPilot enabled for that server, the Control Plane can trigger an autonomous remediation attempt — for example, restarting a crashed application daemon.

Uptime History

The Health Dashboard displays an uptime timeline for each server, showing a color-coded bar for every 24-hour period. This gives you instant visual insight into recurring instability patterns, such as a daily crash that happens at midnight when a backup script runs out of memory.