Comment 1 for bug 1610157

I have checked another scenario: deployed one controller and then added 2 computes. As a result I have had correct /etc/hosts contents on all nodes. It looks like this bug is specific to node's delete process.