Dağıtık veritabanı konfigürasyonlu stok kontrol sistemi

Abstract

Tez, dağıtık veritabam mantığına ? göre hazırlanmıştır. Dağıtık veritabanı sistemleri teknolojisi veritabanı sistemi ve bilgisayar ağı teknolojilerinin birleşiminden oluşur. Bu tip bilgisayar sistemleri, kendi veritabanına sahip olan, kendi programlarını çalıştıran ve aralarındaki iletişim sistemi üzerinden mesaj alışverişi ile haberleşip bağlantı kurarak sonuç üreten işlemciler sistemidir. Kullanıcıların verinin nerede ve nasıl saklandığını bilmesine gerek yoktur, onlar sistemi dağıtık olmayan tek bir sistem olarak bilirler. Tez uygulaması, mantıksal olarak birbiriyle alakalı ve coğrafi olarak dağıtılmış veritabanlan ile çalışan bir firmanın satış/pazarlama, planlama, satın alma, üretim, maliyet ve sevk konularım içeren entegre bir stok kontrol çalışmasıdır. Veritabanlan bir bilgisayar ağı üzerine dağıtılmıştır. Her yerel bölge kendi bilgilerini tutar, fakat gerektiğinde birbirine X.25 TURP AK uzak iletişim ağı ile bağlı olan bu sistemler birbirleriyle bağlantı kurarlar. Tez dört ana bölümden oluşmaktadır: Birinci bölümde; ilişkisel veritabanı yönetim sistemleri, bilgisayar ağları, ISO/OSI referans yapısı, TCP/IP katmanları, internet protokol ve adresleri hakkında genel bilgiler yer almaktadır. İkinci bölümde; dağıtuk veritabanı, dağıtık veritabanı yönetim sistemleri, temel prensipler, dağıtık hareketlerin yönetimi, avantajlar ve dezavantajlar, problemler her yönüyle anlatılmaktadır. Üçüncü bölümde; uygulamanın yazılımında kullanılan progress ilişkisel veritabanı yönetim sistemi yapılan, dosyalan, sınırlamalar, hareketler, kilitler, progress ve TCP/IP ağlan, bağlantılar, güvenlik yönetimi ve kurtarma denetimi anlatıl maktadır. Dördüncü bölümde; proje konusu ve iş akışı, veritabanı ve uygulamalar yeralmaktadır.

This thesis comprises mainly four sections. The first section explains Relational Database Management Systems and Computer Network. The second section deals with Distributed Database Management Systems. The points discussed in this section are what a distributed database is, its advantages and disadvantages, its comparison with central systems, and the twelve rules and problems regarding them. The third section discusses the Progress Relational Database Structures on which this thesis is based, their files and limits, the relationship between Progress and SQL, transactions, locks, multi-database. connections and details with security management and its effects with examples. The last section of the research paper, which is also the project stage, gives information about the subject of the project and workflow, provides explanations of the sections and their distribution, the master files and the special structures used in the project. This section also supplies examples about databases accompanied by their applications. Since the 1960's almost every ten years there have been various improvements in the database. The first data model was a hierchical one and was applied to large systems. Since the 1970's there has been a shift to relational data model and the model has been used not only in large systems but also in smaller systems and personal computers. In the 1980's semantic data modals were introduced. Also the fourth generation languages and logical programming were used over this period. Since the early 1990's combined data models, information display and programming languages, parallel operations, optical memories and distributed data processing have been used. These improvements in the database technology have been continuing at full pace. In order to have an easy access to data and to meet the needs of users as soon as possible, computer firms and distributors use distributed database management systems which utilize advanced concepts and technologies. In case of heavy data processing in a central computer, there is an abnormally high demand for data processing. This, in turn, makes access to data difficult and even pose problems which could cause the entire system to crash. Centralized systems store data in tabels in a central DBMS. Problems often arise in communication lines of centralized systems, making data transmission over distances difficult (Figure 1). O Payroll O Personnel O Stock Figure 1. An ordinary multi-user centralized configuration. In a distributed configuration, data and applications are located in sections independent of each other. Every section has its own data and applications as well as CPU and DBMS. If the system breaks down locally only in one section, this will not affect the other sections (Figure 2). This project has been developed for a pharmaneutzial firm that is involved in planning medicine production and control. The organization has been developed as a distributed database system which are spread over 3 local areas. The areas are Ankara, Levent, Topkapı. The project has been made using a progress relational database in the context of UNIX. The databases were spread over a computer network. Each local site stores its own data, but these systems which are connected to a X.25 TURPAK distant communications network, if necessary, communicate with each other through TCP/IP. This project, which completely deals with inventory control, has the following sections: XI WAN DBMS Local database 2 CPU SITE B SITE C Figures 2. A multi-users distributed configuration. 1- Sales / Marketing 2- Purchasing 3- Planning 4- Raw material / Wrapping storing 5- Product store 6- Production 7- Quality control 8- Cost 9- Delivery store These sections operate in connection with each other. Failure of one of the sections affects the entire system. Acoordingly, malfunction in a section causes the othre sections to operate slowly, and even worse, in a wrong way. Medicine prices are determined and, if necessary, with drawn by the Ministry of Health. The workflow in the project is as follows: xn 1- Clients applies to the Sales/Marketing department with a demand. 2- The Sales / Marketing department conveys the demand to the planning department. 3- The department prepares the production plan if the demand cannot be supplied. 4- A MRP necessary for the production is made. 5- The purchasing department informs the planning deparment of the latest corrections (amendments) made. 6- After all control has been made, the planning department conveys its demand to the purchasing department. 7- The raw materials are checked for quality. 8- The materials checked and passed are ready for production. 9- Workcommands are prepared. 10-The production is started on the workcommands. 1 1-The medicines produced are wrapped and marketed. Local databases, when necessary, can be connected to each other by means of some functions and structures in the progress relational database management systems developed in this project. One of the database connection functions is CONNECT. The point which should be considered in connecting by means of a CONNECT function is that first the database is connected and then the subfunctions are operated. CONNECTED ilac. RUN altprg.p. altprg.p : FOR EACH ilac.stok. DISPLAY stok. END. Starting a server or broker for a database ia distinct from connecting a database. As always, you must start a database server or broker for a database, before any user çan connect it in multi-user mode. Although it is possible to connect to several databases within one CONNECT statement, it is a good idea to connect only one database per CONNECT statement. This is because a connection failure for one database causes a termination of the xni current CONNECT statement, leaving any subsequent database specified with the CONNECT statement unconnected. It is possible to connect to a database during a PROGRESS session using the Progress Data Dictionary.