. "In this case, you would put together a series of server clusters focused on NFS or CIFS that would handle the user queries instead of having the mainframe pinged each time there is a query." . . .