Discovery Hub allows you to gradually replace an existing data warehouse with a new MDW solution, or to use the structure of an existing database as the basis for a new data warehouse. This capability is supported by two Discovery Hub features: the external SQL connection, and the reverse data warehouse function.
Reversing a data warehouse allows you to copy tables and constraints, views, stored procedures, and user defined functions from an existing SQL database and import them into your project as metadata.
These objects will then need to be re-created through the user interface into a new database. They can be re-created in the same database if the original imported objects they are based on are removed.
Note: This feature is only meant to help you get started creating a data warehouse schema and is not a 1-click solution for recreating a data warehouse. You will still need to transform the data, and create data movements in Discovery Hub.
If you want your users to be able to access an existing data warehouse database as you work to rebuild or replace it using Discovery Hub, external SQL connections are a great solution.
The default behavior of this feature is to create views of an existing SQL database in your new Discovery Hub MDW. For this to work, the databases must be located either on the same server, or to linked servers. If the existing data warehouse database is on a different server, you will need to enable data transfer. Data transfer with external SQL connections does not support incremental loading, so this feature may not work smoothly with a very large data warehouse solution located on a different server.
Once you have added tables from your existing data warehouse to your MDW, the tables (or views) will appear in dark gray in the Discovery Hub interface and have very limited capabilities compared to a standard Discovery Hub table. However, they will behave exactly as normal data warehouse tables and views do from your end user’s perspective. They can serve as a source for new SSAS, Tabular, or Qlik models.
Once you have finished building out the ETL infrastructure necessary to populate one of your legacy data warehouse tables in your new Discovery Hub project, you can convert an external SQL connection view into a normal table, and map your staging area tables to it as if you’d created it by hand through the Discovery Hub user interface.