[6.3.0-rc4] Shutdown processing does not complete when VMs take too long

Description: The "Shutdown time-out" value of 60 seconds is too short to let the rc.libvirt script finish.

How to reproduce: Shutdown with a VM that hangs.  This happens when a Windows 10 VM is processing updates.

Expected results: The VM would at least be forced to shutdown and unRAID shutdown processing would continue normally.

Actual results: The VM is shutdown with the equivalent of pulling the plug and stopping services events are not processed.

Other information: The rc.libvirt script is used to shutdown VMs and has a 60 second timer it uses to allow VMs time to shutdown normally.  If the timer runs out, VMs are forced to shutdown.  The "Shutdown time-out" value of 60 seconds is not enough time for the rc.libvirt script to complete causing shutdown processing to be abruptly terminated.

rc.libvirt snippet:

TIMEOUT=${TIMEOUT:-60}
OPTS=${OPTS:-" -f /etc/libvirt/libvirtd.conf -p $PIDFILE "}

check_running_machines() {

  count=0

  for machine in $(/usr/sbin/virsh list --uuid --state-running | grep -v ^$); do
    /usr/sbin/virsh shutdown $machine
  done

  echo -n "Waiting on VMs to shutdown"

  while [ $(/usr/sbin/virsh list --name --state-running | grep -v ^$ | wc -l) -gt "0" ]; do
    if [ "$count" -ge "$TIMEOUT" ]; then
      break
    fi
    echo -n "."
    count=$(expr $count + 1)
    sleep 1
  done

  echo ""

  if [ $(/usr/sbin/virsh list --name --state-running | grep -v ^$ | wc -l) -gt "0" ]; then

    echo "The following VMs are still running, forcing shutdown: $(/usr/sbin/virsh list --state-running --name | grep -v ^$ | paste -sd ',')"
    for machine in $(/usr/sbin/virsh list --state-running --uuid | grep -v ^$); do
      /usr/sbin/virsh destroy $machine
    done

    sleep 2
  fi

}

Suggestion:

- Set default "Shutdown time-out" to 90 seconds.

- Do not let a value of less than 90 seconds be used.