Data Division in system design

We sincerely invite James and SAP consultants to provide guidance. Thank you.
Thanks to the presence of Leader Yang. I am so honnered.

Purpose:
It provides suggestions and reference for unreasonable Data Division in self-developed business systems, and also serves as a research on the design and processing modes of general business systems.
We hope that the system we make will reach the bright realm of reasonable business, clear logic, core business separation, and loose coupling as soon as possible.

Scope of discussion:
Basic Theories and Practices of data classification and its relationship with Design Patterns

Glossary:

preface:
with the deepening of software in enterprise applications, the complexity of business scenario descriptions that users describe directly increases. In the conception and analysis process from business to system, various analysis methodologies provide us with a variety of methods to analyze problems, for example, the UML (mainly usercase and Lane) advocated by the traditional up theory, the user stories advocated in the AP methodology (I do not know much about it), and M $ last year (maybe the year before) the proposed DSL (domain specific languages) should also have its methodology (for vs2005 platform, there are currently few materials available for learning ). On the other hand, analysis methods in traditional software engineering (see software engineering apractitioner's approace) focus on data input and output analysis methods, unable to adapt to the fast response Software Project Process (it seems AP is suitable ). However, is the traditional analysis method useless? The monks think that the traditional method plays an irreplaceable role in data processing, especially in the analysis and design of core business separation and loose coupling. (This can be seen from the SAP system design)

logical division of business systems in enterprise applications:
enterprise applications can be logically divided into business systems and business analysis systems (mostly OLAP ). (Large-scale applications are usually divided logically and physically at the same time ). The business system goes deep into the business process (the department involved in the business), Processes related transactions of enterprise applications, and generates daily business data. The business analysis system processes and generates useful information based on the business system, for enterprises to conduct their own management and monitoring, and adjust their solutions and strategies (Dongdong for leaders ). The data produced by the Business System is raw data, which is called plain data in data mining. The data processed by the business analysis system is often called refined data, is the essence of plain data refined for special purposes. Generally, a general business system includes a simple report function that provides simple data analysis. After a Data Mining Engine is created for a more advanced application, to have a deeper level of analysis capabilities. Theoretically, from data mining, any data has meaningful information under a certain condition, which is called information entropy (Dr at the beginning of school. daniel said, so far it will not be counted. But in fact, the business system cannot keep any data. However, if the business system does not obtain raw data, no matter how hard the analysis system works, it cannot be done. There is a conflict between them. How can we choose how to divide them. This article focuses on the data content division of business systems and some design principles.


data type definition:
common data partitioning methods in enterprise applications, data is defined as three types: system data, master data, and business data biz data. In its system, SAP separates organization data from each other, which is a relatively independent concept. Business data also introduces the concept of business master data, which is also treated differently.

system data:
system data refers to the data required for the normal operation of the system, the lack of this part of data or incomplete data will cause the system to be unable to run or run abnormally, and this part of data will be basically determined at the initial stage of development, it will not change in the future. (System Data: it is the data which R/3 system needs for itself .) the most common is the system configuration file (for example, the system menu structure data, which is generally considered as system data). For Java and DOTNET platforms, XML Data tends to be used as system data.
the system data and business are logically independent from each other. In many business systems, it is easy to confuse personnel data and permission data, as well as the so-called " Code data" (for example: type Configuration of friends and friends.

Organization data:
Organization data can be understood literally. It refers to the data configured by the Organization in the business system. It enters the system at the System Configuration stage and is defined by the customer at the same time. After it enters the system, this part of data rarely changes. (It is a customizing data which is entered in the system when the system is configured and is then rarely changed .)
It is of great significance to divide organization data from system data. In terms of origin, the organization also has the primary data and system data. However, the data of organizational units is different from the general primary data in that the Organization data is "rarely changed". Once the Organization data changes (especially the hierarchy changes), for the business relationship in the entire system, the allocation of accusations of business roles will change accordingly. The size of the impact scope of the change, because of the size of a system, that is, the size of the business scope applied by the system (whether it is a separate material system or an associated MIS, or more closely related business ERP. The data of personnel, roles, permissions, and processes has a major impact, which is reflected in the system business layer. This impact will be devastating in the current situation where many systems are mixed with processes and core services, as well as permissions and core services.

@ Todo
Primary data definition, role, and core position
Foreign Research on primary data processing and mature solutions (including SAP)

@ Todo
The process of business data generation has the significance for subsequent data processing.
Relationship between core business data (primary business data) and non-core business data.
The design principle of business data fetch.

@ Todo
How to fetch data for OLAP and design reasonable business data content