7. Data Center - Data Development: Data System Construction

7.1　 Data system planning 
		 The raw data generated by various systems are stacked together, resulting in very high use cost , This kind of data can only be used in some departments with a very good foundation of data technology , And often 
	 Different names 、 Different caliber issues , As a result, the data of the whole enterprise cannot be really used . The data system of data center is based on the original data of the whole domain , Carry out standard definition 
	 Hierarchical modeling , The final result of data system construction is a complete set 、 standard 、 Accurate data system , It can easily support data application .

		 The middle station data system shall have the following characteristics ：
			1. Cover global data 
				 Centralized data construction , Cover all business process data , The business can always find the required data in the middle office data system .
			2. The structure is clear 
				 Vertical data layering , Horizontal topic domain 、 Business process division , Make the whole hierarchy clear and understandable .
			3. The data are accurate and consistent 
				 Define consistency indicators , Unified name 、 Unify business meaning , Calculate the caliber uniformly , And a professional team is responsible for modeling , Ensure the accuracy and consistency of data .
			4. Performance improvement 
				 Unified planning and design , Choose a reasonable data model , Clearly defined and standardized , And consider usage scenarios , The overall performance is better .
			5. cost reduction 
				 The construction of data system enables data to be shared by business , This avoids the repeated construction of a large number of chimneys , Save calculation 、 Storage and labor costs .
			6. Convenient and easy to use 
				 The general principle of ease of use is that the more you go forward, the more convenient it is to use data directly , Put some complex processing as far as possible , If necessary, make appropriate redundancy treatment . For example, in the data 
			 In the use , Related processing can be carried out in advance through dimension redundancy and fact redundancy , To avoid calculation only when used , Through public calculation 、 The coexistence of details and summary is business 
			 Provide flexibility . The construction of a unified data system gives the business of the whole enterprise the opportunity to use data .


		 In order to make the data system have the above characteristics , Need a systematic data hierarchy , This hierarchical architecture defines the data hierarchy and the modeling construction specification of each layer . data 
	 An architecture is a set of guidelines , In the implementation process, it shall be implemented in strict accordance with the architecture . The architecture of data center involves the following 4 Multiple data tiers ：
		1