I'm playing with gluster on amazon EC2 and recently hit the issue of UUIDs being identical on machines launched from the same Amazon Machine Image. I would like to create a system in which these machines are to a certain extent disposable, and where some sort of automatic discovery and scaling can happen. To that end, I think I would like a random UUID generated on instance launch. How is the UUID generated and can anyone think of any problems with this idea? Is it sufficient to just overwrite /etc/glusterd/glusterd.info with a new random UUID line? Thanks, Brandon
