Comment 2 for bug 1653337

The logs here show intermittent connectivity issues to the fuel master node. At the same time it might have happened that we either have some environmental misconfiguration for these tests, e.g. in amount of RAM allocated to the nodes. As far as I see, there are 5 controller nodes being bootstrapped which means that the node needs at least 24 gigs of ram for controllers and the master node. This could be an issue in case we are running these tests on too tiny jenkins slave nodes.