fix #370: run deployment_bastion_node_cleanup only on the same server uuid

[ioggstream]

This patch should fix the very annoying #370

Before unregistering the node it checks if the server uuid attached to the OS::Heat::DeploymentConfig is the same of the target machine.

As the ip address could be reused in time (eg. if the server is removed instead of just marked unhealthy) I used the server uuid, relying on

/var/lib/cloud/instances/$node_id 

Feedback welcome + Peace, R:

[markllama]

I'm going to see if I can try this this week. it may be that this also solves the problem with a stack delete hanging.

[ioggstream]

@markllama great. Please let me know if you need any kind of support. Feel free to email:robipolli@gmail.com / irc:ioggstream me.