> Worst case is that you can wait 9,999second if the resource is updated just > before the vm stop/start. Sure , I did not say it is perfect ;-) But it is acceptable on average (for me).