Rack aware cluster failure causes higher number of objects

At this point I am fairly sure this is due to deleted records on the restarted servers coming back due to this