> Our disaster recovery scenario, which has been invoked three or four > times, involves running a VE via NFS until a local copy exists, and it > works pretty well. However, they've never run like that for more than > about a day. > can you please develop a little bit more this setup?