The database persistence will store the data for the repository in a relational database, storing each node as a single row. The node and its properties are serialized into a binary form using BSON, and this binary representation is stored in a BLOB column in the underlying database. While most property values are stored with the node, some property are too large to be included in the serialized BSON representation: doing so would increase the deserialization time. For example, most binary values are likely to be larger than the node's BSON representation, and so they are stored separately in the repository's binary store. Also, the repository can be configured so that string values larger than some threshold size are also stored as binary values in the binary store. (And because binary values are immutable, they are identified by the SHA-1 hash of the value; thus if the same content is stored on different nodes as a binary property, the binary store will actually only store one copy.)

You need to configure both kinds of stores, but you certainly can use the same database for both.