We might benefit from splitting the clustering packages into two
separate groups for high-availability and high-performance computing.
This will help to keep the bugs a bit cleaner so each group of folks
will have a clearer idea of what they need to do.