Configurations of this kind, where end-users have access to a dataset that is derived from the master database (and which may be a sub-set of that master, tailored to give access only to data needed by the individual user on the 'need to know' principle) are commonly called "replicas" (sometimes "clones"). Once a replica has been created it operates independently of the master, and its data content will need to be "syncronised" with the master from time to time (by file transfer or on-line).