Conceptual data warehouse in the data warehouse is an integrated, integrated, relatively stable, reflecting a data collection of historical changes, is used to support management decisions. 1. Oriented: Unlike the operational database, the theme is an abstract concept, which refers to the key aspects of the user who cares about the decision of the data warehouse. Reflecting the side of the business process, not as separated as an operating database. 2. Integrated: Operating databases are usually related to certain specific applications, and the database is often independent of each other, they are heterogeneous. The data warehouse is integrated with the original dispersion data, and the data type transformation is performed, and the inconsistency in the data is eliminated. 3. Relative stability: The data warehouse is stable. Because of the use of enterprise analysis decisions, it will be reserved for a long time, and the reference query will be given. And he is relatively stable because the data warehouse should save historical information to reflect the changes of a matter, as the original information of analysis decisions, so the data warehouse must maintain historical information, so he is relatively stable. Write Once, Read Many Times.4. Reflecting History: To support analysis decisions, the processes have been summarized by the past development history, thus making reasonable forecasts for future.
Data Warehouse Environment 1.Olap: Data Extraction, Transformation, and Loading 2.OLAP: Online Analytical Processing Engine3.dss: Decision Support System 4. Customer Analysis and Report Tool 5. Other Data Collection and Data Output Tools
The architecture of the data warehouse can be divided into four levels: data source, data management and storage, OLAP servers, and front-end tools. 1. Data Source: He is the foundation of the data warehouse, the bottom of the data warehouse architecture, is the data source of data warehouse. Includes information of each business handling subsystem. 2. Data Management and Storage Layer: It is the core of the data warehouse. How efficient management data is different from the main criteria for the operational database. Follow the subject management data, the aggregate data is stored in the multidimensional database. 3.OLAP Server: Effective integration of data and organizes by multidimensional model. 4. Front-end tool: mainly includes various report tools, query tools, data mining tools, and the like.
The process of ETL data into the data warehouse is generally taken three processes extracted, transformed and loaded, which are referred to as ETL. Generally use existing tools to implement ETL process. 1.e: Data extraction. Data cannot be modified while data extraction. The file format that can be extracted is: Database object, such as the table can be exported from the source system throughout. For example, MS SQL Server 2000 BCP "Select * from northwind..customers" Queryout "D: Temp.txt"? -C -p -u "sa" -p "sa". The extraction process is dynamically extracted, that is, the target changes, which does not affect the efficiency. One implementation is to increase timestamp; another method is to describe in different tables, respectively. 2.T: Data Transformation. The data will be transferred from one system to another system. The order is: Source System -> Staging Database -> Data Warehouse -> Data Mart. There are three ways: A.FLAT Files B. Distinguish the operation c. Use the switch partition. 3.L: Data input
other
Data warehouse does not have to use a database to implement
