What are the database management systems

Database system

A database system is a software system that manages the data in a database. A database system can be described by three components: database management system (DBMS), logical structure of the data and physical structure of the data. In addition to the system components, the relevant environment of the DB system is entered. It consists of the users, the external storage media and the operating system. The requirements of the users can be summarized in the functions of adding, changing, inserting and deleting data. There are two ways in which users can contact the database. On the one hand, they can access the data contained in the database with the help of application programs. In this case it is necessary to establish a connection between the programming language used and the DBMS. This is done by introducing database commands into the programming language that are understood by the DBMS. The programming language serves as the host for the database commands. It is therefore also called "hostlanguage". The advantage of such systems is that they can be programmed efficiently and are able to process complex logical structures. On the other hand, users without programming knowledge can send queries directly to the DBMS. You use your own database language. The summary of all database commands is called database manipulation language (DML). In addition to the DML, the DBMS also contains the data description language (DDL). With their help, the scheme and the sub-schemes are described. Since many manufacturers claim that their data management system is a database system, the term s should be explained in more detail to distinguish it. A database system must meet the following criteria:
a) There is a database management system with the following tasks: Administration of the database structure Administration of the data descriptions Administration of the data and physical storage including the establishment of the logical relationships Coordination of access to the data backup functions (AFTERIMAGE, BEFOREIMAGE, LOGGING) Data protection (password administration, etc. on physical and logical levels )
b) There is a schema that contains the data descriptions and the relationships (structure) between them. In addition, several application-specific excerpts from this data description (sub-schemes) are stored. These are administered by the database management and made available to the corresponding applications.
c) With the elementary database commands, an execution is carried out directly on the physical level without the inclusion of file management functions (area / page concept, summary of records of the same type is not mandatory). The advantages of using a database system are: extensive freedom from data protection, better data protection options than with file management, program independence, parallel processing options, multiple access, good restart options, uniform naming of data elements. A disadvantage is the lower performance compared to conventional file management. The use of database systems is spreading more and more. Most of the systems offered (see ISISSoftwareReport) are based on the network data model proposed by the CODASYL group (Data Base Task Group). A new data model, the relation model, is currently even more widespread in theory than in practice, although there are already database systems for this on the market. The following table shows a list of the most important database systems (source: ISISSoftwareReport, system programs,
2. 4/1982)

is created by combining several files, taking into account the logical dependencies between them. A database system is a software system for managing the stored data under a common interface. It is described by three components: • Database management system (DBMS), • Schema / subschema (logical data view), • Structure of the data on the storage media (physical data view). The DBMS is a software package for the administration of the complex data and memory structures with the help of the operating system and the access procedures defined for the files. The task of the DBMS is to guarantee the required logical data structure (data view). The schema and the sub-schemas are defined by a data description language (DDL) (data dictionary). The data manipulation language (DML) contained in the DBMS enables the user to access and manipulate the data as required. The relevant environment of a database system consists of users, external storage media and the operating system. End users can also send queries to a database without any programming knowledge. For this purpose, the DBMS has its own database language (query language). On the other hand, it is also possible for users to access the data contained in the database via an application program. In this case it is necessary to establish a connection between the programming language used and the DBMS. For this purpose, database commands are inserted into the programming language; in other words, the programming language serves as the "host language" for the database commands. However, programming skills are required for this second form. Further functions of a database system are: • Data protection: There are access restrictions for users with regard to files, records, fields, field contents or also with regard to the possibility of reading or changing data. • Data backup: All status changes are automatically checked (before and after images). • Data efficiency: Competing resource allocations and deadlocks are prevented. By using database systems, the following advantages can be achieved compared to file organization: • Extensive freedom from redundancy of the data stock, • Data backup enables restart in the event of system errors and errors in application programs, • Data protection, • Data are understood as a separate organizational element and are not part of the application programs , organizational regulation of parallel access. • On the other hand, these diverse functions are bought with a high computer load (overhead) compared to conventional file processing.      

A database system is understood to be a database in connection with a database management system for the consistent and secure storage, management and use of this data (database management system). Database systems form the essential basis of business information systems, be it in the context of so-called online transaction processing (OLTP) for handling operational business processes or in terms of so-called online analytical processing (OLAP) in the context of decision support systems.

The architecture of database systems can be described using the so-called three-level chemical architecture (standardized in the ANSI / X3 / SPARC model); The structure is shown using three different perspectives or levels of abstraction, which are described below.

The logical schema describes the structure of relevant data regardless of the type of physical storage and specific application perspectives. The logical schema is derived from a conceptual schema that encompasses the overall semantic view of the data (for example in a comprehensive, integrated company data model). In the representation of such schemes, graphic description methods such as the entity elationship model or formal data definition languages ​​are used in particular. The basic entities (object types) of the world segment to be modeled, their properties (attributes) and their relationships are primarily described (possibly refined as general associations, aggregations, generalizations / specializations, etc.). This also includes the mapping of simple consistency conditions (for example, such that an object of a certain object type must have a certain relationship with exactly one object of another object type). For the representation of a simple excerpt from the data of a library, compare the Entity Relationship Model.

While the term logical schema is often used synonymously with the conceptual schema, the logical schema in the narrower sense is aligned with the underlying data model. This is mostly about the relational data model, in which data structures are only represented by means of a central construct, the relation. Based on this, the relational database systems dominate in operational practice (a relation is then also referred to as a table). Relational database systems are well suited for mapping relatively simply structured data. A more recent approach is object-oriented (data) modeling. In the internal scheme, the physical or technical description of the structure of stored data is defined (in particular the file organization including access options). This physical data organization is essentially independent of the logical data model (physical data independence). External schemas describe specific sections (views) of the logical data model. This makes the relevant data views available to various applications (for example, an application to support purchasing logistics, the supplier data, but no customer data). This enables the previously close dependency between specific applications and the storage of data to be separated (logical data independence). In this context, the definition of access rights to the database must also be regulated, also in order to meet the requirements of data protection. A database system has a central data catalog (data dictionary) to store the (meta) information about the database that is mapped in the various schemes. The role of a database administrator includes the functions of setting up and maintaining such a data catalog. Database management systems store, manage and process the data; in particular, they enable transactions to be carried out on the database (for example, the secure execution of a booking). A transaction is understood to be a sequence of elementary database operations that are to be viewed as an atomic unit with regard to certain integrity conditions. In this context, the essential requirements for database management systems are summarized in the so-called ACID principle (atomarity, consistency, isolation, durability). Database management systems should guarantee that transactions are always carried out atomically (completely or not at all), consistently (according to defined consistency conditions), regardless of the simultaneous execution of other transactions, and permanently. This ensures data security and consistency. To access the data, database management systems provide at least programming language interfaces, classifiable in a data definition language for defining the data schemes, a data manipulation language for changing access to the data and a data query language for reading data access. SQL (Structured Query Language) is the dominant relational database language in practice. Extensive standardization and implementation in practice that is consistent at least with regard to the language core makes it possible in principle to change the database management system without having to change the applications that access the database system.

1. Historical starting situation Apart from more or less trivial systems such as a pocket calculator, every information system needs permanent data in order to fulfill the tasks for which it was developed. The fundamental importance of permanent data in information systems is that their existence is an indispensable prerequisite for any kind of memory in an information system. This importance as well as the fact that the management of permanent data is on the one hand relatively complex, but on the other hand can be designed largely independent of the content, already in the early years of operational information processing led to the fact that subsystems of information systems were isolated and subsequently reused The task was to manage persistent data. The first subsystems of this type were file systems, but since they were primarily designed for the management of all kinds of data within the framework of operating systems, they were too general to adequately support the user-friendly management of application-specific data. In particular, within the framework of file systems it is not possible to control the contextual relationships between different files on the system side, and the simultaneous use of databases by several users is not sufficiently supported. Both deficits are impressively remedied by database systems according to the current state of the art.
2. Basic structure On closer inspection, a database system is composed of the fundamentally application-neutral database management system and the application-specific database, the database management system being merely a system framework that does not have any usable functionality without a database. A database management system typically supports exactly one, and occasionally also several data models. The database of a database system consists of a database scheme and the data structured according to this database scheme. The reason for this conceptual separation is the fact that a database changes over time in that new data is inserted into it, data contained in it are changed and data is deleted from it. What remains stable over the long term compared to this short-term variability of the data is the structure of the data. The database schema thus includes the long-term stable aspects of a database, while the associated data can be changed in the short term.
3. Abstraction levels and data independence Since the end of the 1970s, in connection with databases, the distinction between three abstraction levels has proven to be useful and established. Each of these three levels of abstraction, the internal, the conceptual and the external level, represents its own way of looking at both the database schema and the data. On the internal level (including the physical level), the data in the database is perceived in its entirety and in the database schema describes how they are actually stored in the form of data records and how they are arranged on the storage medium. At the external level, the data in the database is presented as it is required by the individual groups of users or application programs. This means that on the external level there is a separate database schema for each such group, which only describes the part of the database required by the respective group in the representation required by the respective group. At the conceptual level, the entirety of the data in the database is presented in a way that is as neutral as possible (in contrast to the external level) and is also described in the database schema. Only a few data structures that are very simple compared to the internal level are used for this representation. In this way, the conceptual level hides a large part of the complexity existing on the internal level, which results from the fact that ideally there should be the possibility of an optimal organization on the internal level for every type and every form of use of data. The way the data is presented at the external level and at the conceptual level differs little from one another; in an external level database schema, in comparison to the conceptual level database schema, only data is typically omitted primarily. In this way, some of the complexity on the external level that can result from the diversity of data on the conceptual level is eliminated. In addition to the aforementioned reduction in complexity, the main advantage of differentiating between the three levels of abstraction is the resulting data independence.
4. Data manipulation A database management system must provide instruments with the help of which access to the content of a database is possible and changes can be made to both the content and the structure of the database. First of all, a differentiation must be made according to who or what such access is supposed to take place, taking into account different stakeholder groups such as so-called naive users of the database who have no IT background whatsoever, occasional users who expect a certain amount of training application programmers and the programs they develop, and finally database administrators.Against this background, it may come as a surprise that at least modern database systems based on the relational data model largely manage with a single database language, namely SQL. SQL was originally aimed at occasional users in the specialist departments of companies who were supposed to be able to access data in databases without the support of the IT department. In the meantime, the suitability of the language for application programming has also been proven, with the database language being coupled with a conventional programming language in one of various possible ways. The administration activity can also be largely based on SQL, whereby the data that is accessed in this case are for the most part not the application data in the database, but the associated metadata. Database languages ​​such as SQL are not suitable for naive users, such as the counter staff of residents' registration offices, transport companies or banks. Rather, when accessing a database, this group of people is limited to calling up prepared programs and only making very specific, strictly monitored entries in masks or graphical user interfaces.
5. Competitive use Database systems that play a role in our economic life are typically multi-user systems. This means that a very large number of users work with the system at the same time. As long as these users only have read access to the database, this is not a problem, since the users cannot interfere with each other. However, this changes at the moment in which even a user or a program operating on the database tries to make changes to the data. In this case, the database management system must take precautions to ensure that changes made by a user only become visible to other users when they have been made completely and irrevocably. Furthermore, on the DBMS side, it must be ensured that the quasi-simultaneous execution of changes by several users does not result in data states that could not have arisen if the changes had been executed consecutively. Problems such as these have led to the fact that the change in data in databases is typically broken down into processing units, called transactions, of very short duration, with very high demands being made on their reliability for these transactions. The concept of the transaction also plays an important role in restarting a database system after a fault. In this case, the content of the database must be brought into a state that is characterized by the fact that the transactions in progress at the time the fault occurred have all either been properly completed or have left no effect on the data; In addition, it must be clear which of the transactions must be executed again. Even without going into technical details in this regard, it should be clear that such requirements can only be met with a great deal of effort by the database management system and are therefore largely responsible for the complexity that is typical for the software of a database management system. If, on the other hand, one considers the transfer of funds from one account to another as an example of transactions, then it becomes clear that many database system operators would not be able to conduct their business using computers without such high reliability requirements .
6. Outlook In the past, database technology has mainly concentrated on alphanumeric data as it occurs in our economic life. For some time and especially under the impression of the possibilities of the Internet it can be observed that increased focus of research in the field of database systems on the storage and use of other types of data such as texts, time series, geographic data, sound, image - and video data is directed. It turned out that the functionality of traditional database management systems can only be used to a limited extent for these new types of data. Although it is in principle possible to “saddle up” the corresponding additional functionality on existing systems, this leads to software that is difficult to understand and therefore ultimately difficult to maintain. Against this background, a new generation of database management systems can be expected in the future, in which texts, alphanumeric data, program code, data streams etc. are integrated into a uniform architecture. Note For the related areas of knowledge, see Business Intelligence, Business Networking, Controlling, Controlling Information Systems, Data Warehouse, Electronic Government, ERP Systems (Enterprise Resource Planning Systems), Management Information Systems (MIS), Business Informatics, Basics, Knowledge Management, —3 Workflow management.

Literature: Abiteboul, S. et al .: The Lowell Database Research Self-Assessment, Communications of the ACM, vol. 48, no. 5 (2005), pp. 111-118. Date, CJ: An Introduction to Database Systems, 6th ed., Reading, MA, 1995. Garcia-Molina, H., Ullman, JD, Widom, J. : Database Systems: The Complete Book, Up-per Saddle River, NJ, 2002. Kemper, A., Eickler, A .: Database Systems: An Introduction, 5th Edition, Munich, 2003. Lockemann, PC, Dittrich, KR: Architektur von Database Systems, Heidelberg, 2003. Maier, D .: The Theory of Relational Databases, Rockville, MD, 1983. Ramakrishnan, R., Gehrke, J .: Database Management Systems, 2nd ed., Boston, MA, 2000. Ullman, JD: Principles of Database and Knowledge Base Systems, vol.1: Classical Database Systems, Rockville, MD, 1988. Ullman, JD: Principles of Database and Knowledge Base Systems, vol.2: The New Technologies, Rockville, MD, 1989. Ullman, JD, Widom, J .: A First Cours in Database Systems, Upper Saddle River, NJ, 1997.

Previous technical term: database schema | Next technical term: database systems

Report this article to the editors as incorrect & mark it for editing