Disk Space Watch

Hourly disk usage check on the host running InTouch. Notifies when any mount exceeds a configured percentage.

What it does

The monitor's check runs a bash one-liner that walks every real disk mount (filtering out tmpfs/devtmpfs/squashfs), prints a per-mount status line, and emits the literal token ALERT when at least one mount is over the threshold. The when: arm uses contains ALERT to gate notification.

Quiet by default — if all mounts are under threshold, the else: arm matches and nothing fires.

Edit before installing

In monitor.yaml:

• THRESHOLD_PCT=85 — change to your alert percentage.
• The df filter excludes tmpfs/devtmpfs/squashfs. Add -x <fs-type> for any additional pseudo-filesystems your host has.
• For REMOTE host monitoring, swap the runtimeenv check for ssh to the target machine. The same bash logic works over SSH.

Bindings used

• {{check.output}} — full stdout of the bash check, including the per-mount status lines.

Tools used

• runtimeenv — core InTouch task type that runs the inline bash script under the named runtime environment (bash).

Pattern showcase

{{check.output}} contains ALERT is the string-marker pattern — a cheap, no-parse way to gate a notification on the result of a free-form shell script. Works for any check that can emit a known sentinel string ("ALERT", "FAIL", "OVER", etc.) when something is wrong.

Previous form

Replaces the disk-space-alert job (deleted 2026-05-19). The job version fired its notification unconditionally with the same body; this monitor stays quiet on healthy days.