2010/04/15 15:44:20

DBMS

The Database Management System (DMS) - the specialized complex of programs intended for the convenient and effective organization, control and administration of databases. As the DBMS structural form any of the models existing today can be used. Relational DBMS or network DBMS can be an example of such model.

The solutions DBMS directory and projects is available on TAdviser.

Content

Outsourcing of DBMS
- 2020: In what cases it is necessary to transfer to outsourcing of DBMS and the business application
What requirements should modern DBMS meet?
The market of DBMS in Russia
Classification
File systems
DBMS of large computers
Desktop DBMS
Products
History
Links
See Also

The company is larger, the longer it exists in the market, the more data accumulate in its archives. And, according to realities of today, all information is stored electronically. According to the research Aberdeen Group, three years ago in the large companies volumes of the stored information increased by 32% annually. Today the business intelligence operates with already multiterabyte amounts of data, and storages clearly move to cloud platforms.

However what information, invaluable for business, would not be stored on servers of the company, it should correspond to several criteria as if data are entered haphazardly and are stored in different formats – the advantage of the resources spent for this purpose is very doubtful.

DBMS represents a set of programs which in total manage the organization, data storage in a DB. In general such systems are classified depending on their data structure and their types. DBMS accepts requests of application programs and instructs the operating system for transfer of the relevant information. New categories of data, can be added to a DB without violation of the existing scheme. The organizations can use one type of DBMS for implementation of daily transactions, and then place necessary information by other machine which works with other management system more suitable for accidental requests and the analysis. Servers of backup of databases, as a rule, are multiple processor systems with the large volume of OZU and large disk RAIDs. DBMS actually is heart of the majority of applications for work with a DB.

Outsourcing of DBMS

2020: In what cases it is necessary to transfer to outsourcing of DBMS and the business application

In April, 2020 the director of the center of technical consulting RDTECH Pavel Shmelyov listed the main pluses of outsourcing and told why the companies give maintenance of own DBMS on the party. Read more here.

What requirements should modern DBMS meet?

Information storage should be replenished constantly with new data according to a rhythm of life of the company, same also formation of new categories of accounting concerns.

At the same time to enter information any new employee should have an opportunity. The same concerns also service of this infrastructure from the system administrator, formation of new data samplings from analysts with the different tolerance level to information and with different profiles of the analyzed data.

It should be noted that the amount of information increases not only in volume, but also is qualitative. As a result there is a need of simultaneous operation with it several experts. Besides, there is a possibility of involvement of specialized experts for accomplishment of difficult procedures of data analysis (Data mining Intelligent data analysis). Today not only for formation of future strategy, but also for accomplishment of routine tasks the predictive analytics which for formation of a right vector of development uses objective, but not subjective data has the increasing value.

In addition the single database simplifies reportings to both separate divisions of the company, and all company in general for submission of documents in the state instances or their granting for acquaintance to external commissions of experts. The uniform base allows to update documentation relating to the normative reference information (NRI) in the automatic mode. In addition general database allows to operate with information in the peculiar unified form, doing procedures of information analysis by more uniform within the company.

One more plus from uniform approach to information storage is much simpler procedure of reservation, decrease in a downtime after failure, data security provision in the dispersion plan of access rights, more transparent procedure of migration on new versions program and the hardware.

The market of DBMS in Russia

Основная статья: DBMS (market of Russia)

The Russian DBMS are in quite difficult situation today — developments of the large companies squeeze out the Russian systems from the market. However, still there are several DBMS which continue to remain afloat due to implementation in government institutions.

Classification

Depending on architecture of creation of a management system for DBMS bases can be subdivided into the following types:

File systems

Let's imagine that there is some information medium of a certain capacity, the device for reading record on this carrier in the mode of random access and application programs which use the specific carrier for input-output of information in an external bulk memory. In this case, each application program should know where and in what place necessary data are stored. As it is more application programs, than information media, several application programs can use one drive. What will occur if one of application programs needs to dozapisat the data on a disk? There can be an imposing: a situation in which data of one program will be rewritten by other program. An important step in development of information systems was transition to use of centralized control systems of files. In terms of the application program, the file – the referred to as area of an external bulk memory to which it is possible to write data, and from which it is possible to read out them. In order that there was an opportunity consider information from any area of an external bulk memory it is necessary to know a name of this sector (file name), the size of the area and its physical location. The file control system performs the following functions:

distribution of an external bulk memory;
display has files in the corresponding addresses in an external bulk memory;
ensuring access to data.

Consideration of features of implementation of separate file control systems is beyond this subject. At the let stage rather know that application programs see the file as linear sequence of records and can execute over it a number of transactions. The main transactions of a sfaylama in SUF:

create the file (a certain type and the size)
open earlier created file
read a certain record from the file
change record
add entry in the end of the file

DBMS of large computers

This stage of development is connected with the organization of databases by big IBM 360/370, EU-EVM machines and different models of Hewlett Packard company. In that case information was stored in an external bulk memory of the central computer. The tasks started generally in batch mode were users of databases actually. The interactive mode of access was provided using console terminals which had no own computing resources (the processor, RAM, an external bulk memory) and served only as input/output devices for the central computer. Programs of access to a DB were written in different programming languages and started as normal numerical programs. Features of this stage:

All DBMS are based on powerful multiprogram OS (Unix, etc.).
Work with the centralized DB in the mode of distributed access is supported. Control functions by resource allocation are performed by the operating system.
The languages of low data manipulation focused on navigation data access methods are supported. The significant role is assigned to administration of data.
Serious works on justification and formalization of a relational data model are carried out. The first system (System R) implementing ideology of a relational data model was created.
Theoretical works on query tuning and management of distributed access to the centralized DB are carried out, the concept of transaction was entered.
Big flow of publications on all questions of the theory of a DB. Results of scientific research actively are implemented into commercial DBMS.
There are first languages of the high level for work with a relational data model (SQL), however there are no standards for these languages.

Desktop DBMS

Computers became closer and more available to each user. The awe of ordinary users before unclear and difficult programming languages disappeared. There was a set of the programs intended for work of unprepared users. And clear transactions of copying of files and transfer of information from one computer on another, printout of texts, tables and other documents became simple. System programmers were pushed the background. Each user could feel the complete owner of this powerful and convenient device allowing to automate many aspects of own activity. And, of course, it affected also work with databases. Newly appeared DBMS allowed to store considerable information volumes, they had a user-friendly interface for filling, the built-in means for generation of different reports. These programs allowed to automate many accounting functions which were conducted manually earlier. Permanent reduction of prices of personal computers made it software available not only for the organizations and firms, but also for certain users.

Computers became the tool for maintaining documentation and own accounting functions. All this played both a positive, and negative role in the field of development of databases. The seeming simplicity and availability of personal computers and their software generated a set of laymans. It was much created systems phony companies which did not answer laws of development and interrelation of real objects. However availability of personal computers forced users from many knowledge domains which did not apply ADP equipment in the activity earlier, to address them. And demand for the developed convenient programs of data processing forced software providers to deliver all new systems which it is accepted to call desktop DBMS. The considerable competition among suppliers forced to improve these configurations, offering new opportunities, improving the interface and high-speed performance of systems, reducing their cost. Existence in the market of a large number of DBMS performing similar functions demanded development of methods of export import of data for these systems and opening of the data storage formats. But also during this period there were fans who illogically developed own DBMS, using standard programming languages. It was the deadlock option because further development showed that it was much more difficult to transfer data from non-standard formats to new DBMS, and in certain cases required such labor costs that it would be easier to develop everything again, but data all the same should be transferred to new more perspective DBMS. And it was too result of underestimation of those of function which DBMS had to execute. Features of this stage following:

Standardization of high-level languages of data manipulation (development and deployment of SQL92 standard in all DBMS).

All DBMS were expected creation of a DB generally with exclusive access. And it is clear. The personal computer, it was not connected to network, and the database on it was created for work of one user. Consecutive work of several users, for example, at first the operator who entered accounting documents, and then the chief accountant who defined the postings corresponding to primary documents was in rare instances assumed.

The majority of DBMS had developed and a user-friendly interface. In the majority there was an interactive mode of work with a DB both within the description of a DB, and within design of requests. Besides, the majority of DBMS offered the developed and convenient tools for development of ready applications without programming.

In all desktop DBMS only the external presentation layer of a relational model, i.e. only an external tabular style of data structures was supported.

In the presence of high-level languages of data manipulation like relational algebra and SQL in desktop DBMS low-level languages at the level of separate lines of tables were supported.

In desktop DBMS there were no means of support of referential and structural integrity of the database. These functions had to run applications, however the scarcity of development tools of applications sometimes did not allow to make it, and in this case these functions had to be performed by the user, demanding from him additional control at input and change of information which is stored in a DB.

Existence of an exclusive operation mode actually led to degeneration of functions of administration of a DB.

Rather modest requirements to the hardware from desktop DBMS. Quite operable applications developed, for example, on Clipper worked at PC 286. In principle, it is even difficult to call them full-fledged DBMS. Bright representatives of this family — Dbase DBMS which were very widely used until recently (DbaseIII+, DbaseIV), FoxPro, Clipper, Paradox.

Products

The solutions DBMS directory and projects is available on TAdviser.

History

Databases were used in ADP equipment from time immemorial. In the first computers two types of external devices – magnetic tapes and magnetic drums were used. Capacity of magnetic tapes was rather high. Devices for reading record magnetic tapes provided sequential access to data. For reading information which was in the middle or the end of the magnetic tape it was necessary to read all previous section at first. Was a consequence of it extremely poor performance of input-output operations of data in an external bulk memory. Magnetic drums gave the chance of random access, but had limited scope of the stored information. Certainly to speak about any data management system in an external bulk memory, at that moment it was not necessary. Each application program which needed to store data in an external bulk memory itself defined arrangement of each block on the magnetic tape. The application program also undertook functions of information exchange between RAM and devices of an external bulk memory using software and hardware tools of the low level.

Such operation mode does not allow or very much complicates support on one carrier of several archives of for a long time stored information. Besides, each application program had to solve problems of naming of parts of data and structurization in an external bulk memory. History DB actually began with the advent of magnetic disks. Such devices of an external bulk memory had significantly bigger capacity, than the magnetic tape and reels and also provided many times over the high speed of access in the mode of an arbitrary access. Unlike modern management systems which can be applied to the most different databases the vast majority of earlier developed DBMS were closely connected with the user base to increase work speed, though to the detriment of flexibility. Originally DBMS were applied only in the large organizations with the strong hardware support necessary for work with large volumes of data.

The increase in productivity of personal computers provoked development of DBMS as separate class. To the middle of the 60th years of the last century there was already a large number of commercial DBMS. Interest in databases increased more and more so this sphere needed standardization. The author of the complex Integrated Data Store database Charles Bachman organized a target group of DTG (Data Base Task Group) for the approval of features and the organization of the DB standards within CODASYL - groups which were responsible for standardization of the COBOL programming language. In 1971 the code of statements and notes which was called Approach of CODASYL was submitted, and after a while there were first successful commercial products made taking into account notes of the above-mentioned working group. In 1968 also the IBM company which provided own gpod DBMS the name IMS was noted.

Actually this product represented compilation of utilities which were used with System/360 systems on shuttles Apollo. The solution was developed according to kotspeta of CODASYL, but at the same time the strict hierarchy was applied to structurization of data. In turn in CODASYL option network DBMS was taken for basis. Both options, meanwhile, were accepted by community later as classical options of the organization of work of DBMS, and Charles Bachmann in 1973 got Turing's award for work the Programmer as the navigator. In 1970 the employee of IBM company Edgar Kodd working in one of departments of San Jose (USA) in which were engaged in storage development of systems wrote a number of articles concerning the DBMS navigation models. Having become interested in a question he developed and stated several innovation approaches concerning the optimal organization of management systems of a DB. Kodd's work made the significant contribution to development of DBMS and is the valid founder of the theory of relational databases. 1981 E.F. Kodd created a relational data model and applied to it transactions of relational algebra.