Which is the purpose of VM Component Protection (VMCP)?

Disconnected virtual NICs are handled by network-level HA features or guest OS recovery, not VMCP, which is scoped solely to storage failures.

VMCP monitors the storage fabric and triggers automated responses when a datastore experiences a Permanent Device Loss (PDL), where the storage array signals the device is unrecoverable, or All Paths Down (APD), where no I/O paths remain available. Administrators can configure VMCP to issue warnings or restart affected VMs depending on the failure type.

Restarting failed applications within a guest is the role of vSphere Application Monitoring or guest-level watchdog processes, not VMCP.

Virus detection and response inside a virtual machine is a guest security function handled by endpoint protection tools, not a vSphere HA storage protection feature.