Code review comment for lp://staging/~hopem/charms/precise/cinder/lp1226823

Fair enough, i'm sure I saw that value used in the past but your method based on the number of OSDs makes more sense. I'll adjust that. I'm not clear on what we want to do for the replication count though. What we have right now is a count of 2 and Adam mentions single node testing which would still be broken by a count of two since we replicate by host by default. I will remove the replication count set for now since it is not currenlty breaking anything but we really need the ability to modify this since production deployments will likely want a replication count of 3.