The database approach has some very characteristic features which are discussed in detail below:
A database system allows several users to access the database concurrently. Answering different questions from different users with the same (base) data is a central aspect of an information system. Such concurrent use of data increases the economy of a system. An example for concurrent use is the travel database of a bigger travel agency. The employees of different branches can access the database concurrently and book journeys for their clients. Each travel agent sees on his interface if there are still seats available for a specific journey or if it is already fully booked.
Structured and Described Data
A fundamental feature of the database approach is that the database systems does not only contain the data but also the complete definition and description of these data. These descriptions are basically details about the extent, the structure, the type and the format of all data and, additionally, the relationship between the data. This kind of stored data is called metadata (“data about data”).
Separation of Data and Applications
As described in the feature structured data the structure of a database is described through metadata which is also stored in the database. An application software does not need any knowledge about the physical data storage like encoding, format, storage place, etc. It only communicates with the management system f a database (DBMS) via a standardised interface with the help of a standardised language like SQL. The access to the data and the metadata is entirely done by the DBMS. In this way all the applications can be totally seperated from the data. Therefore database internal reorganisations or improvement of efficiency do not have any influence on the application software.
Data integrity is a byword for the quality and the reliability of the data of a database system. In a broader sense data integrity includes also the protection of the database from unauthorised access (confidentiality) and unauthorised changes. Data reflect facts of the real world. database.
A transaction is a bundle of actions which are done within a database to bring it from one consistent state to a new consistent state. In between the data are inevitable inconsistent. A transaction is atomic what means that it cannot be divided up any further. Within a transaction all or none of the actions need to be carried out. Doing only a part of the actions would lead to an inconsistent database state. One example of a transaction is the transfer of an amount of money from one bank account to another. The debit of the money from one account and the credit of it to another account makes together a consistent transaction. This transaction is also atomic. The debit or credit alone would both lead to an inconsistent state. After finishing the transaction (debit and credit) the changes to both accounts become persistent and the one who gave the money has now less money on his account while the receiver has now a higher balance.
Data persistence means that in a DBMS all data is maintained as long as it is not deleted explicitly. The life span of data needs to be determined directly or indirectly be the user and must not be dependent on system features. Additionally data once stored in a database must not be lost. Changes of a database which are done by a transaction are persistent. When a transaction is finished even a system crash cannot put the data in danger.