Multiple nodes in the current pre-beta cluster were in a bad state (node NotReady). Upon reboot, several nodes came up with CoreOS1235, requiring the flannel fix. Additionally, two nodes (node4, gfs2) would not reboot – exhibiting the same problems as the LoadBalancer on 1/12/2017.

Working with Mike, tried the following:

• Shutdown node4, gfs2. Rename to node-dead
• Detach volumes
• Re-run ansible openstack-provision.yml to see if nodes are correctly re-created