Introduction

Becoming of control systems by databases has coincided on time with significant successes in development of technologies of the distributed calculations and parallel processing. In result there were control systems of the distributed databases. These systems become dominating tools for creation of appendices of intensive data processing.
In the environment distributed  database the decision of the questions connected to increase of volume of databases or needs of processing becomes simpler. Thus seldom there is a necessity for serious reorganization of system; expansion of opportunities is usually reached due to addition of processor capacities or memory.
In an ideal distributed database possesses property of linear expansibility and linear acceleration. Linear expansibility is understood as preservation of the same performance level at increase in the size of a database and simultaneous proportional increase in processor capacity and a memory size. Linear acceleration means, that with escalating processor capacity and a memory size at preservation of the former size of a database productivity proportionally grows. And at expansion of system the minimal reorganization of an existing database is required.
In view of a parity price / productivity for microprocessors and workstations to make system of several small computers more economically, than to realize her by one big machine equivalent on capacity. Set commercial distributed database function on minicomputers and workstations owing to more favourable parity price / productivity. The technologies based on application of workstations, have received so a wide circulation due to that the majority commercial database are capable to work within the framework of local networks where basically and workstations are used. Development distributed database intended for global networks WAN, can lead to to increase of a role of mainframes. On the other hand, distributed database the future generations, most likely, will support the hierarchical network structures consisting from кластеров within the limits of which computers cooperate on the basis of a local network, and кластеры incorporate among themselves by means of high-speed highways.
All these reasons have led to to rapid development of technologies of the distributed  database databases of information systems. However there are unresolved some complicated questions connected to the given theme. One of these questions - dynamic processes in the distributed databases.
In the given work principles of functioning of the distributed databases are considered, imitating model database is developed. With the help of the received model the dynamic processes proceeding in database are investigated.

                                                                  Object and object of research

Object of research is the distributed database  database of computer information systems.
Object of research is influence of dynamic processes on functioning of the distributed database.
Distributed database database database is a set of logically interconnected databases distributed database in a computer network.


In other words, systems of the distributed databases will consist of a set of the units connected together by a communication network, in which:
1) Each unit possesses own systems of databases;
2) Units work in coordination, therefore the user can get access to the data on any unit of a network as though all data are on it(him) with own unit.

.
The control system of the distributed database  is a program system which provides management of the distributed  database so that its distribution was transparent for users.
It is possible to add these definitions if to consider also various characteristics database and database In 1981г K.Date has published the rules for the distributed database databases [1.] these are below resulted 12 rules.
1. Local autonomy. The local data should be under local possession and management, including functions of safety, integrity, data presentation in memory. Exception can be a situation when restrictions of integrity cover the data several at units or when management of the distributed transaction is carried out by some external unit.
2. Any concrete service should not be assigned to any specially allocated central unit. Observance of this rule, i.e. a principle децентрализированности functions database , allows to avoid bottlenecks.
3. A continuity of functioning. The system should not stop in case of need additions of new unit or removal in the distributed environment of some data, changes of definition of the metadata and even (that is rather difficult) realization of transition to new version database on separate unit.
4. Independence of a site. Users and not about are obliged to know appendices about where the data physically settle down.
5. Independence of a fragmentation. Fragments (named also sections) the data should be supported and processed by means database  that users or appendices could know in general nothing about it. Moreover, database should be able to bypass at processing inquiries the fragments which are not having to them the attitude.
6. Independence of duplicating. The same principles of independence and a transparency concern and to the mechanism of duplicating.
7. The distributed processing inquiries. Processing of inquiries should be made by the distributed image.
8. Management of the distributed transactions. It is necessary to distribute mechanisms of management to the distributed databases transactions and managements of simultaneous access. This problem includes revealing and the sanction of impasses, interruptions after time and intervals, fixing and recoil of the distributed transactions, and also a number of other questions.
9. Independence of the equipment. Same software database should be carried out on various hardware platforms and function in system as the equal in rights partner. In practice to reach  it it is extremely difficult, as many suppliers support set of platforms. This restriction is overcome with the help of model of multi-grocery environments.
10. Independence of operational systems. This problem is closely connected with previous, and  also is authorized similarly.
11. Independence of a network. Units can be connected among themselves to the help of set of various network and communication means. The m the Multilevel model inherent in many modern information systems (for example,  model OSI, model TCP/IP, levels SNA and DECnet), provides the decision of this problem not only in database environment, but also for information                                                            systems in general
12. Independence from database Local database  should have in an opportunity to participate in functioning database.
2 Analysis of technologies of organization database
The database is physically distributed on units of computer information system by means of a fragmentation and   of the data [2, 3, 4].
2.1 A fragmentation of the data
The attitudes belonging to a relational database, can be  on horizontal or vertical sections.
The horizontal fragmentation is realized by means of operation of selection which directs each train of the attitude to one of sections, being guided by a predicate of a fragmentation. For example, for attitude Employee  the fragmentation is possible according to territorial distribution of workplaces of employees.
Then inquiry " receive the information on employees of the company " can to be formulated so:

SELECT * FROM employee@donetsk,
employee@kiev



SELECT employee.emp_id,
emp_name,
salary
FROM employee@donetsk,
employee@kiev
ORDER BY emp_id

Due to a fragmentation the data come nearer to a place of them most a heavy use, that potentially reduces expenses for transfers; the sizes of the attitudes participating in the user inquiries decrease also. However practically to achieve acceleration of performance of the inquiries touching  the attitudes, very difficultly. The basic problem will consist in sharp expansion of space of search of variants of performance of inquiries with which should work  inquiries.

The second way of distribution of the data   means creation of duplicates of the data.is a set of various physical copies of some object of a database  for which synchronization  with some "main" copy is supported.

Theoretically values of all data in the reprinted objects should automatically and be synchronized immediately with each other. (In practice this rule is weakened usually a little.) In some systems of a copy are used exclusively in a mode of reading and are updated according to the set schedule. In other environments updating of separate values in copies is supposed, and these changes are distributed according to procedures of planning and coordination  various models of duplicating are shown.
At  fragments of the data are duplicated in view of demand for access to them. It is useful, if access to the same data is necessary from the appendices which are carried out on different units. In that case, from the point of view of economy of expenses, it will be more effective to support copies of the data on all units, than continuously to send the data between units.

The basic problem the data is that updating of any logic object should be distributed to all stored copies of this object. Difficulties arise that some unit containing the given object, can be inaccessible (for example, because of crash of system or the given unit) at the moment of updating. In that case obvious strategy of immediate distribution  on all copies can appear unacceptable as it is supposed, that updating of transaction will be failed, if one of copies will be inaccessible in a present situation.