ARTICLE: Database Management: Architecture and Data Models

The architecture is the backbone of the design of the DBMS. The three categories of the design of the database are the decentralized, centralized and the hierarchical. Also, how the architecture of the DBMS can be seen are the single as well as multi-tier. The number of the steps of the design can be anything because the modules are independent and they can be altered, replaced, and modified. N number of modules are the dividends of the n-tier architecture of the whole system.


If we talk about the 1-tier system architecture, the only entity where there is direct access to the users to the data is the DBMS. The users can even use the data directly. In addition to this, they can even make the changes in the system. Any changes made here will modify the system of the database management directly. But, the end users are not provided with the handy tools. The programmers and the database designers tend to go for the single-tier architecture.


For the working of the data in the format of the 2-tier system, there is a requirement of an application with which the users can access the DBMS. In the 2-tier architecture of the system, the programmers have access to the modern database management system with the help of an application. The other name of this tier is the application tier, and it is entirely independent of the designing, programming, and the operation of the system.


The divisions in the 3-tier architecture are based on the user complexity and how the users would present the data in the database. The designers and the programmers prefer this type of the architecture of the system for the designing of the DBMS.

The 3-tier architecture of the DBMS consists of the presentation tier, application tier and the database tier.

  • Database or Data Tier:
    In this place, the database has to get the company of the query processing languages. In this architecture, the relations of the data and the constraints at this level have to be defined.
  • Application (Middle) Tier:
    There are the application server and the programs which are used to access the database of the system. If you want to have an abstract view of the database, then the application tier performs the best in this case. Only the administrators know about the database. The end-users have no idea of it. They only see the application. In the prior tier of the database architecture of the system, the tier has no information of any other user further than the application tier. The layer which acts as the mediator between the database and the end-user is the application layer as it sits in the middle.
  • User (Presentation) Tier:
    In this tier, the end-users operate on this. But they are unaware of the existence of the database beyond this layer. The users can have the multiple views of the database, and the application provides it. The applications generate all the views which reside in the application tier.

The tier which is highly modifiable is the multi-tier. It is because it consists of all the components which can be changed and independently.

There are several data schemas and data independence which are related to the DBMS.


The models which are used for defining the modelling of the logical structure of a database. They are the necessary entities which are used for the introduction of the abstraction of the data in the DBMS. The approach which is used for the connection of the data with each other includes in the data models. Also, the process of the data stored inside the system is constituted in the data models as well.

The flat data models, the very first data model, in which all the user data has to keep on the same plane. The data models in the earlier times were prone to the introduction of the updated anomalies and the duplication of the data due to their scientific approach.


The data model, the entity-relationship model, includes the real-world entities and their relationships. This data model begins creating the entity set, general attributes, and the constraints when the formulation of the real-world scenario into the database model is to start.
It is the database model which is used for giving the conceptual design of the database. The elements on which the entity-relationship model is based are as follows-

  1. Entities and the Related Attributes
  2. Relationships Among Entities


According to the entity-relationship model, the real-world entities which have attributes as their properties are called as the entity. In this, the attribute which is defined by the different set of values is called the domain. For instance, in the ER model, the entity would be a student if we talk about the school system. Moreover, the values of the attributes can be anything from the name, age and the classes etc.


The relationship of the entities is the logical association of one entity with the other. There are various ways in which the relationships are mapped. The function which is used for defining the number of association between the two entities are the mapping cardinalities.

These are as follows-

  • One to one
  • One to many
  • Many to one
  • Many to many


Relationship model in the most popular data model in the DBMS. In comparison to the other data models, it is the one which is the most scientific. The logic used in this model is the first-order predicate logic. Also, it defines the tables as the n-ary relations.

The main points or we can say the highlights of this model are the-

  • The relations of the data model can be normalized.
  • The area in which the data is stored in the tables is called the relations.
  • The values which are saved in the normalized version of the data model are the atomic values.
  • Every row of the relation of the model constitutes the unique model.
  • The source of the values for each column in the relation is from the same domain.