1. forms an isolation layer between the business system and the data warehouse.
General data warehouse application systems have very complex data sources, which are stored in different geographical locations, different databases and different applications. Extracting data from these business systems is not an easy task. Therefore, using ODS to store the data directly extracted from the business system is basically consistent with the business system in terms of data structure and logical relationship between data, so the complexity of data conversion is greatly reduced in the extraction process, and the main concerns are the interface of data extraction, data size and extraction method.
2. Transfer the function of querying some business system details.
Before the establishment of data warehouse, a large number of reports and analysis were directly supported by business systems, which brought considerable pressure to the operation of business systems in the process of generating some complex reports. The data of ODS is consistent with the business system in granularity and organization, so the query of reports and detailed data originally generated by the business system can naturally be carried out from ODS, which reduces the query pressure of the business system.
3. Complete some functions that cannot be completed in the data warehouse.
Generally speaking, in the data warehouse architecture using ODS, the data stored in DW layer is summarized, and the detailed data generated by each transaction is not stored. However, in some special applications, it may be necessary to query the detailed data of transactions, so it is necessary to transfer the detailed data query function to ODS, and the data model of ODS is stored in a theme-oriented way, which can easily support the query functions such as multidimensional analysis.
In the architecture of data warehouse application system without ODS layer, the granularity of data stored in data warehouse is determined according to needs, but generally speaking, the most detailed business data also needs to be retained, which is actually equivalent to ODS. But unlike ODS, the detailed data at this time is not "current and changing" data, but "historical and unchanging" data.