<div>
  (Optional, a whole number) What to do when automatic host selection finds <b>no host that the VM can be
  created on at the moment</b>: how many <b>seconds to wait</b> for one before giving up. This covers
  every such situation, which the build log names when it happens:
  <ul>
    <li>the <i>free resource limits</i> rule out every host (busy hosts: expected to pass);</li>
    <li>no host has the VM's memory size free right now, if <i>Require enough free RAM</i> is on (busy hosts);</li>
    <li>all hosts are disconnected, in maintenance mode or not among the candidate hosts;</li>
    <li>no host has enough cores or RAM for the VM, if those requirements are on;</li>
    <li>no candidate host reports CPU/memory usage statistics.</li>
  </ul>
  The value means:
  <ul>
    <li><tt>0</tt> (the default): do not wait; give up at once.</li>
    <li>a positive number: look at the hosts again every 15 seconds for up to that long.</li>
    <li><tt>-1</tt>: wait for as long as it takes (the build can still be aborted).</li>
  </ul>
  <p>
  Giving up means: <b>failing</b> the clone/deploy if it was the free resource limits that ruled out
  every host (vCenter must not place the VM regardless of them); in all other situations, leaving the
  placement to vCenter, as was always done. The build log says which of these applies, and whether it
  will wait, right when the situation is found. A clone or deploy step, or a template, can set its own
  wait time. A waiting build holds its place in the queue and its connection to vCenter, so a limited
  wait is advisable on busy farms.
  </p>
</div>