@george1421 host lookup from the server is different, it doesn’t use port 80. It uses some ICMP query that you’d probably know more about than me. It’s not a standard ping, and is faster and returns more than response/fail.
Remember @Joe-Gill, the logs you need to view should be those of the “master” nodes. That doesn’t mean you shouldn’t see anything, but I found earlier today (23-JUL-2016) that the “halt” is because the node is not a master. I’ve fixed it in the RC-2 and all appears to be working properly now.