Post parsed host watchdog metrics to monitoring platform

Your watchdog timeout events arrive unstructured, leaving engineers without per-host metrics for triage and SLA tracking. This sends parsed host metric points to your monitoring platform so on-call can identify affected hosts before the next incident window.