Allow for a more robust configurations
*) A/A groups.
*) "Not" dependency for resources that should avoid being on the same node.
*) Dependency between clusters or groups.
Tim Cerling commented
It would be helpful to explain what you mean. I'm sure it made all the sense in the world to you, but your terse descriptions leave a lot of question marks.
For example, anti-affinity already exists for resources that should avoid being on the same node. How is that not sufficient for your second example? And, I'm not sure at all what you mean by Active/Active groups. Active/Active clusters have been available since day one, but not all applications can run A/A. So is this a cluster issue or something else? Please elaborate.