Watchdog monitors and protects vCenter Server services. If any services fail, Watchdog attempts to restart them. If it cannot restart the service because of a host failure, vSphere HA restarts the virtual machine (VM) running the service on a new host.
Watchdog can provide better availability by using vCenter Server processes (PID Watchdog) or the vCenter Server API (API Watchdog).
You can use the service start command to start PID Watchdog and the service stop command to stop it. PID Watchdog monitors only services that are running. After the service is stopped, PID Watchdog does not monitor it. PID Watchdog detects only that a process with the correct executable is in the process table. It does not determine if the process is ready to service requests.
Starting with vSphere 6.x, a Python daemon called API Watchdog checks the status of APIs for the VPXD service. If the APIs are not running, API Watchdog attempts to restart the service two times. If that still does not solve the issue, API Watchdog then reboots the VM.
API Watchdog starts running immediately after deployment of the vCenter Server Appliance. On vCenter Server for Windows, however, you must reboot vCenter Server once before API Watchdog starts working.
API Watchdog generates support bundles before a service restart and also before a VM reboot and these support bundles are stored in C:\ProgramData\VMware\vCenterServer\data\core\*.tgz on vCenter Server for Windows and in /storage/core/*.tgz for vCenter Server Appliance.