Well, I'm not sure that Virtualmin has a built=in way to do exactly that.
There were a few threads in regards to redundant setups, clustering, and the like. For example:
http://www.virtualmin.com/forums/general-discussion/rendundancy-clustering-and-failovers.html http://www.virtualmin.com/forums/virtualmin/clustering-virtualmin.htmlPerhaps some of those offer solutions that would be interesting to you?
I guess the overall point you'll read in those, though, is that there isn't one best way to handle redundancy. What you'll probably be doing is build the clustering/redundancy -- then configure Virtualmin to take advantage of it.
-Eric