One way to make this kind of interesting is to address how you'd do a reasonable zeroconf effort given that you need to boot 1m+ machines. We've booted 4400*250 VMs on a machine at sandia, and, let me tell you, it was a pain. It is amusing to watch the programs traverse million line /etc/hosts file -- for a while.
So we're working on approaches in which files -- e.g. /etc/hosts -- are replace with programs. Key is that there can be no configuration files that are linearly related in size to the number of nodes -- which kills dhcpd, hosts, and so on. 9P servers make this sort of thing bearable. Just FYI. ron