Description of problem:

rgmanager doesn't support running multiple rgmanager groups on CMAN
clusters.  Adding this option would allow the subdivision of a cluster
with a large amount of nodes into smaller fail-over clusters, all of
which would still be bound to the constraints of the larger clusters.

I don't think this is a useful option, given that the nodes are still
bound by the main cluster's operation.  This means that when the large
cluster's quorum is dissolved that all of the rgmanager groups
(regardless of their membership) will cease operation.