With CNS, volumes are not created when one of the nodes/gluster pods is down

Issue

1) A brick is unavailable due to hardware issue or the node or pod is not running

2) When OCP creates a new VM, it creates a new Volume (REPL X 3)

3) When Gluster creates a new Volume, and one of the bricks is bad, it errors out (expected behaviour)

4) Glusters error causes OCP to error