Olya Orda

Faculty:	Computer Science and Technology (CST)
Department:	Computer monitoring systems (CMR)
Speciality:	"Computer ecological-economic monitoring"
Theme of Master's Work:	"The monitoring system to improve management faculty"
Scientific Supervisor:	Ph.D.( in Engineering), dean of the CST Faculty Alexander Anoprienko

Abstract

In this article analyzed the processes of information systems evolution. By evolution we mean the modification of the information system in accordance with new technical requirements [1]. As an example of this evolution will be considered in the expansion of the information system of the Donetsk National Technical University, which provides adding subsystem of performance monitoring. It should be noted that the evolution of any system of closely connected with the process of refactoring. According to [2] the process of refactoring is a change in the internal structure of the system without affecting the external behavior and to facilitate its further modification. Though information systems are heterogeneous, because it includes relational and object-oriented component [3], you should analyze the evolution of each of its components individually. In particular, there are distinguish refactoring object-oriented systems [2, 4] and refactoring databases [5] in the literature. It should be noted that the work will increasingly focus on refactoring databases. The reason for this is the current architecture of information system DonNTU. In it isn`t checked the level of the server software, which would isolate the client software from direct access to the database. Therefore, all the client modules are interacting with database for receiving data and data placement. Such an approach to building architecture has a negative impact to the structure of databases that, in particular, manifested in the absence of qualitative data normalization, as well as in their redundancy and duplication [6].

It should be noted that the use of a simple architecture has led to the emergence of a strong connection between the client software modules and database structure. The point is that the development of many client modules that make up the information system used by means of rapid development. This led to the fact that the structure of the database is often expanded and changed in accordance with the requirements of specific client module, without regard to what any overall strategy. This, in turn, leads to the inability to use most forms of refactoring proposed in [5]. The reason lies in the fact that due to the abundance of client modules, which directly interact with a variety of ways databases can not be clearly divided the internal structure and external representation. But in this case, as follows from the consideration of the definition, the process of refactoring can not be executed. Therefore, the studies performed will be directed at identifying the mechanisms that would allow integration of new custom module in the existing system in such a way as to clearly delineate the boundary of their interaction. That is, the purpose of this study is to find approaches to expand the information system that would ensure the lowest possible connection of a new component to the already existing ones. Achieving this goal will avoid complicating the core information system, in this case the central database, while adding to it new functionality. For example, one of the mechanisms of exclusion can serve as a SQL-representation [7]. In fact, this virtual tables that are built based on real tables and / or based on other representations. They allow, in particular, to provide the level of abstraction between the data necessary application and the actual data stored in the database [7]. Submissions are an important mechanism of abstraction, which is provided by most DBSs

Using representations has several advantages, firstly, that if the internal structure of the base change, the interface provided by the representations may remain stable. Secondly, a reporting mechanism is supported by the DBMS that we can expect fairly high performance when using them. Third, if gradually transfer client modules on the use of representations, then later you can get rid of the direct connection client software to the internal structure of the database. Another, much more complicated option is to create a level of the server software [3]. This level can provide the necessary client software interface for accessing data. In addition, this level may provide some useful functions, such as access control to data. In addition, a server-level allows the use of thin clients [3]. That is, the use of a layer of the server software can provide clients with Web-based interface, which is an important factor for modern systems.

Expected to continue to study in detail the interaction of the two described mechanisms. This is, firstly, will make it possible to create a methodology for determining the actions that must be met for the integration of new components into existing system. Secondly, it will identify approaches to the gradual transformation of the system so that it meets current requirements, which address both the internal structure and ways of interacting with users. Previously, you can determine that the process of integrating the module into the existing system will consist of the following steps. First, you need to determine what data will be used module, and provide these data in an existing database. Next, determine what data will be added to the base of an integrable module. For existing data are concepts that will provide better structure for the new data are added to new tables. In this case, should document the expansion base. In addition, if new data is large enough, you can use the additional database to avoid the uncontrolled growth of the existing. New data should depend only on the representations, but not from the original data. That is, the level of representation should ensure that the isolation integrable subsystem-level data.

The next step is to develop server software. For this system of monitoring learning achievement is the software must handle user requests that come through it provides a network interface.

In particular, most queries for the monitoring system will concern the analysis of existing data. In particular, this is analysis of student`sevaluations for a certain period or analysis of certain subjects results. Consequently, most of the time the software will read requests from the database, and these requests will often involve some preliminary data processing. This leads to the necessity for further analysis of another mechanism that has attained a qualitative development in the known databases, namely, the mechanism of stored procedures. This mechanism allows you to perform database-level preprocessing and data preparation. Key advantages of using this mechanism are possible, firstly, to avoid transmission of redundant data, and secondly, to process data faster. So, in the qualification Master's work is expected to perform a study that deals with the search method expansion of existing information systems, in order to add these new features. This study will include analysis of mechanisms for provision of modern database for solving the integration and analysis of contemporary approaches to the construction of the server software. The aim is to establish procedures that will expand the information system so that, at least not degrade its current internal structure. Also, the method should be used to determine priorities for system development. Testing of the proposed methodology will be implemented through the integration of monitoring subsystem performance. The integration will be carried out as follows: the first phase will be designed independent database containing a set of test data. This database will be used in the debug version of the system. However, it has another important purpose - it will formalize those interfaces that are needed to connect the new subsystem to an existing system. In other words, this test database would accurately determine the requirements on the part of the subsystem to the main database. For the design of the database will be used by well-known standard IDEF1X [6]. The second step is the creation of software. As noted earlier, this software is an implementation of Web-service. It should be noted that a number of problems, such as object-relational mapping data on the boundary of the database and the software will be performed by well-known tools platform Java (in particular, to address this technology will be used Hibernate). Initial debugging of the project will be executed regardless of the database information system DonNTU. After the debugging will be done substitution test database as a mechanism to access the main database, for example, representations. The browser will wrap up testing integrable subsystem. Described process of new components development of the information system is part of the work performed in the qualification Master's studies.

Literature

1. Буч Г. Объектно-ориентированный анализ и проектирование с примерами приложений / Буч. Г, Максимчук Р., Энгл М., Янг Б., Коналлен Д., Хьюстон К. – М.: ООО "И.Д. Вильямс", 2008. – 720 с.

2. Фаулер М. Рефакторинг: улучшение существующего кода / Фаулер М. – СПб.: Символ-Плюс, 2007. – 432 с.

3. Фаулер М. Архитектура корпоративных программных приложений / Фаулер М. – М.: Издательский дом "Вильямс", 2006. – 544 с.

4. Кериевски Д. Рефакторинг с использованием шаблонов / Кериевски Д. – М.: ООО "И.Д. Вильямс", 2006 – 400 с.

5. Эмблер С. Рефакторинг баз данных: эволюционное проектирование / Эмблер С., Садаладж П. – М.: "И.Д. Вильямс", 2007. – 672 с.

6. Кренке Д. Теория и практика построения баз данных 9-е изд. / Кренке Д. – СПб: Питер, 2005 – 859 с.

7. Буч Г. Язык UML. Руководство пользователя 2-е изд. / Буч Г., Рамбо Д., Якобсон И. – М.: ДМК Пресс, 2007. – 496 с.

8. Вьейра Р. SQL Server 2000. Программирование в 2 ч. Том 1. / Вьейра Р. – М.: БИНОМ. Лаборатория знаний, 2004. – 735 с.

9. Вьейра Р. SQL Server 2000. Программирование в 2 ч. Том 2. / Вьейра Р. – М.: БИНОМ. Лаборатория знаний, 2004. – 735 с.

10. Василенко А.Ю. Разработка веб-сервиса мониторинга успеваемости студентов /Василенко А.Ю. Автореферат. http://masters.donntu.ru/2009/fvti/vasilenko/diss/index.htm

Note

When writing this abstract the master’s qualification work is not completed. Date of final completion of work: December, 1, 2010. Full text of the work and materials on a work theme can be received from the author or his scientific supervisor after that date.