So-called document stores (also known as document databases) are very popular. MongoDB is probably the most popular representative and is particularly valued for its ease of learning and broad support of programming languages and frameworks. With the new Oracle Database API for MongoDB, MongoDB's Wire Protocol is now "understood" so that MongoDB applications, tools and drivers can connect directly to an Oracle Autonomous Database.
Different approaches: tables and collections
Document store or (object-)relational database: how do the two approaches differ? In the case of (object-)relational databases, schema design plays an important role: objects of the application (e.g. customer or order ) are normalized and thus mapped to argentina telegram screening tables and columns of the database, whereby an application object (entity) often requires several tables. Using schema design not only means good (as complete as possible) knowledge of the attributes and relationships of all application objects, but also knowledge of the relational model (e.g. normal forms) and also SQL knowledge in order to create the database objects (tables). This relational schema determines exactly how the application objects are later stored, modified or loaded. A developer therefore usually needs a minimum of SQL knowledge.
On the other hand, it can also be done without the database as a starting point. Application objects can be stored as a unit (also called Document ) in a NoSQL document store. In the case of MongoDB, the object's attributes are converted to JSON (more precisely BSON) and stored in a collection . The collection is only identified by a name; there is usually no schema that defines or restricts the structure of the JSON data. Each JSON document is assigned a unique value (ID) (automatically or by the user) that can be used to find the document again. A document store is therefore also a key-value store . In addition, documents can also be selected based on search criteria - for example, to find all customers in a city.