Crank is build on Jodi
Java Oracle Data Integrator, Jodi is a successor of code used by Global Business Unit Architecture of Big Red. It uses an intermediate format of textual specifications (XML), to specify transformations e.g. mappings that Jodi #1 code transforms into Oracle Data Integrator mappings, packages, scenarios and loadplans. This intermediate format is ideal for serializing the Extract Transform and Load code to transform a Data Lake ( .csv, .zip, Json, XML, DB ) into a Data Vault Model.
Crank #1 the Data Vault Generator
Adding a source systems (file or DB based) in the extraction layer and creating Primary and Foreign Keys in a disabled state and Crank will generate the Data Vault 2.0 model for you. The underlying Autonomous Data Warehouse artifacts like tables, materialized views, and partitioned tables are generated. The ETL is generated in the intermediate format (XML) so each model may be serialized and stored in Version Control. This intermediate format is input for Jodi to generate the ETL.
Hubs, Satellite, Links everything is generated.
How are we different than other solutions?
* Performance, parallel load plans, partition exchange, utilizes resource manager plans
* Can be (re) generated with different physical properties (hints, ODI KM options)
* Can be stored in VCS
* Runs on Oracle Cloud Infrastructure or On Premises
* Business Vault Creation is on its way