No, DB_FILE (or to be more specific the BerkeleyDB database engine) handles storage in a way that searching for and rewriting of data items are not expensive operations. That is one of the reasons complex database engines are used instead of simply storing data sequentially on disk