Upon changes to node groups that targets different nodes than those were initially, config map for the old groupand NRTs for the old target nodes are not cleaned up.  

    all 

    always

    1. cluster with 2 MCPs for workers with machine count > 0: mcp-1, mcp-2
    2. configure NROP CR with mcp-1; CM and NRTs (following RTEs) will be cretaed for the targetted nodes by mcp-1 
    3. update NROP CR to point to a mcp-2
    

An easy way to reproduce is also:
    1. configure NROP with node group that targets at least 1 node
    2. delete the NROP CR
-> old CM and NRTs are kept

    1. new NRTs are created for the missing nodes as expected
    2. old CM is not deleted 
    3. old NRTs are not deleted

     1. new NRTs are created for the missing nodes as expected
     2. old CM and NRTs should be cleaned up 

    workaround is to delete old (dangling) objects manually