DWH Modernization
With new data platforms, companies can technically update their aging data warehouses. However, this new elastic capability alone does not solve the problem of the slow rate of adaptation of DWHs. In data-driven enterprises, changes are the rule rather than the exception. This is exactly where automation of development activities is needed to integrate new content much faster using best practices.
A coordinated collection of data integration patterns and MetaKraftwerk enables you to create and deploy data integration functionality in an automated way.
A coordinated collection of data integration patterns and MetaKraftwerk enables you to create and deploy data integration functionality in an automated way.
Data Vault (Core)
Loading of hub, satellite and link tables, use for Raw Vault as well as Business Vault.
- Delivery-Modes: FULL, DELTA, REJECT
- Built-in reject handling and rejection of rejected data in reject tables
- Reference check / Foreign surrogate keys
- Duplicate detection
- Multiple-State-Change-Handling
- Non-Chronological-State-Change-Handling
- Custom-Logic-Frame
Transactional Links (Core)
Processing of transactional links (T_LNK), including reject handling. For transactional data (transaction data, mass data) this is a more efficient way of data storage. In addition, the pattern enables advantages in accessing data, since everything is stored compactly in a table.
Dimensional
(Data Mart)
Loading of dimension tables with data from the Data Vault. The arrangement of the dimensions as star and snowflake schema is possible. The pattern makes the historization, which was established in the Data Vault, evaluable (versioning). Complex Data Vault models are converted into optimized, dimensional data structures (Star/Snowflake) for reporting. These take the historization in the core into account. As-what views (master data at the time of the transaction) and as-is views (current status) are available for users.
Facts (Data Mart)
Loading of fact tables with data from the Data Vault, linking with dimensions. For this purpose there are two pointers to dimensions each (L0-ID => As-Is, Versions-Id => As-Was).
Questions about patterns?
In a personal demo, we would be happy to show you the entire functionality of the patterns and how they can bring a decisive advantage to your DWH modernization project.