Architecture of dbms

See in this article some concepts related to data modeling, with emphasis on a DBMS architectures.

Following a series of 'Database Modeling "this is the article related to" Architecture of a DBMS, "which shows in a graphic way the levels of architecture and sample data communications with other databases.

Architecture of a DBMS

Data Modeling

A data model is used to describe the structure "logic" and "physics" of a database. Relationships, data types and constraints are known as the structure or level, dividing into 2 types:

  • High Level - we call conceptual data model or Entity-Relationship model, its main concept is a projection of the data that gets closest to the vision that the user has data.
  • Low - known as physical data model, is what provides a detailed view yet of how the data are aramazenados computer.


When using the term "description" of the database, as we understand the call of "schema of a database" that is specified for a project database.


Instances are formed when a data is saved in the database for a certain time they formed these database instances, being changed every time a change in the database is performed. DBMS ensures that all instances satisfying the schema of the database, respecting its structure and its constraints.

In a DBMS architecture has as main objective, separate user applications of physical data that are divided from the diagrams below:

  • Level domestic or internal schema - uses a data model that shows the physical storage structure of the database, the details of the saved data and access paths.
  • Level conceptual or conceptual scheme - performs a full description of the structure of the database but did not offer details of the data stored in the database.
  • Level or external layout view - describes the views of the database to a group of users that shows which users have access to this database.

Figure 1. Architecture of a DBMS Layers

Data Independence

We can say that is allowed to perform schema changes or level of a database without changing a higher level. Recalling that only those levels or regimens are shown in Figure 1. Below is representing two types of data independence:

  • Logical data independence: change only the conceptual level, without any change in the external or in user applications.
  • Physical Data Independence: change the internally without having to change the conceptual level, externally or user applications

Languages ​​for Data Manipulation

The use of language DDL (Data Definition Language - Data Definition Language) is defined by the conceptual and internal level. When there is a separation of internal and conceptual that is not absorbed a clear view of the user, the DBMS has to compile the DDL action, with the action, permission to execute the statements identified by their descriptions of schemes / levels where you will store them in the DBMS catalog.

Where this detachment has an understanding of the user uses the language SDL (Storage Definition Language - Storage Definition Language) for specifying schema / internally.

By the end, when the DBMS uses the architecture with three levels, the language used is VDL (Vision Definition Language) that serves to deini

Category: Architecture

Similar articles: