Description of problem:
Initial ask is for Ceph block storage only.
Next is Ceph object store. (will need to file a separate bug for it).
Later it can be extended to glusterFS, and other software-define block storage. (yet another bug later)

This is to allow installer to add a node to a storage group.
This includes installing operating system, and all networks setup.
For Ceph block storage it need to include triggering ICE to add the configured node to Ceph cluster, including installing Ceph libraries, rebuilding a keyring, and updating all Ceph cluster and client nodes with new keyring. 

For node deletion consists of two parts.
One is for OSP installer removal of a node under its control.
The second is specific to the storage that is run on storage node.
For Ceph it is equivalent to a failed node, with data redistribution between remaining nodes of ceph cluster, keyring rebuilding and updating keyring on client nodes. But since node to be removed is available it is possible to expedite data redistribution movement.
In addition, an attempt to remove a node from a cluster of size 3 or less is prohibited and returns an error.
UI must be provided for node removal (this is node type independent.)

For node replacement, Installer should provide a way to specify which node in storage cluster should be replaced, it could be down already, or still active.
But new node should attain identity of replaced node, including its network connections and its IP addresses. Then ICE need to "re-install" ceph on new node.

Version-Release number of selected component (if applicable):
N/A

How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info: