Data warehouse architecture10/19/2023 ![]() The DDS contains a complete set of fact tables and the dimension tables. Unlike an NDS + DDS architecture, in an ODS + DDS architecture you have only one DDS. In the ODS + DDS architecture, the DDS is the master data store. Second, it is able to load data into several DDSs. First, it integrates data from several source systems. The NDS is in third normal relational form or higher. This architecture is similar to the single DDS architecture, but it has a normalized data store in front of the DDS. In NDS + DDS data flow architecture, there are three data stores: stage, NDS, and DDS. ![]() The advantage of combining them into one package is to have more control over the timing of when the data is written to and retrieved from the stage. In Figure 2-4, the stage ETL, the DDS ETL, and the data quality processes are combined into one ETL. Data architecture where the stage ETL and DDS ETL are combined The data in the DDS can also be uploaded into multidimensional databases, such as SQL Server Analysis Services, and then accessed by the users via OLAP and data mining applications.įigure 2-4. In Figure 2-3, the “Control + Audit” box contains the ETL control system and audit, the data warehouse applications, such as business intelligence (BI) reports, read the data in the DDS and bring the data to the users. The data flow architecture shows how the data flows through the data stores within a data warehouse.įigure 2-3. A data flow architecture is one of the first things you need to decide when building a data warehouse system because the data flow architecture determines what components need to be built and therefore affects the project plan and costs. In figure 2-2 shows a data flow architecture complete with the control system, the metadata, and the components of data quality processes. A data flow architecture with control system, metadata, and data quality process DDS ETL retrieves data from ODS and loads it into DDS.įigure 2-2. ODS ETL retrieves data from a stage and loads it into ODS. A stage ETL retrieves data from the source systems and loads it into a stage. In Figure 2-1 there are three ETL packages between the four data stores. A data flow architecture with a stage, ODS, DDS, and MDB
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |