So in doing some testing, I pulled the plug on my node where the hosted engine was
running. Rough timing was about 3.5 minutes before the portal was available again.
I searched around first, but could not find if there was any way to speed of the detection
time in order to reboot the hosted engine quicker.
Right now I am only testing this and will add in VM's later, which I understand should
reboot a lot quicker.