while the topic of replication is on, i'd like to ask if anyone has
implimented networked block level mirroring with eNBD
http://www.it.uc3m.es/~ptb/nbd/ for postgres. specifically i'm speaking of
replication for the purposes of high availability, but i'm also interested
in query performance scaling via slave read-only instances.
since i last looked at NDB, they've tacked on the ``e'' - looks like the
project is active although still young.