How about adding the ability to configure a particular bucket for a given schema as opposed to bucket per papercut instance?