Probe management

One-click updates

When a new probe version ships you'll see it on the Probes page (and get a notification). Click Update — the probe downloads the new version, verifies its checksum, installs dependencies, and restarts itself. If anything fails it keeps running the old version and reports why.

• Previous version is backed up on the probe machine automatically
• No remote desktop, no file copying, no site visit

Watchdog supervision

• Windows: a Scheduled Task checks every 3 minutes and relaunches the probe if it stopped — covering reboots, crashes and interrupted updates. It installs itself; no console windows appear.
• Linux: the systemd service restarts automatically.
• Double-starts are harmless — a second probe instance exits cleanly before touching anything.

Probe-offline alerts

If a probe goes 15 minutes without a heartbeat you get a probe-offline alert through your normal channels — because a dead probe means paused monitoring, and silence shouldn't look like health. When it reconnects, the alert auto-resolves and a recovery notice follows.

Code integrity attestation

The probe ships as packed code and reports a hash of what it's actually running on every heartbeat. The cloud compares it against the released build and shows verified / modified on the Probes page — so a tampered or corrupted probe is visible immediately.

Buffering

During internet outages the probe keeps polling and buffers results locally, uploading when connectivity returns.