Method to update table column definitions and use in multiple datasets without semantic modeller?

Question

Hi all!

We are documenting our DW tables and assigning more meaningful labels to the columns for consumption by our users.  The challenge we are encountering is that for each dataset we create, we have to redo all of the column labeling for a given table.

I really hope there is a better approach that allows reuse of the labels rather than making changes in every dataset that uses the table.

Looking forward to your thoughts!

OAldrich · Answer

Thanks @Gianni Ceresa !  We would definitely prefer the governed semantic modeller, however we do not have the Enterprise license which includes that tool.  Ideally, the semantic model would be the best choice!

Without semantic modeller, we also considered creating the datasets with a manual SQL query.  Doing so would allow us to control the labeling in the SQL and simply copy the SQL text from one Dataset to another.  Very cumbersome, but would save some time...

Gianni Ceresa · Answer

Not answering your question directly, but more thinking at something else that could have helped you.

You seem to be doing everything with datasets. Didn't you consider using the semantic model to model your DW tables and expose them to your users? That gives you central control by default, and also comes with simple ways to renamed columns centrally and things like that.

My understanding is that datasets are mostly part of the self-service way of using the tool, which comes with great freedom for users, but less central management tools for governing your data. Because the governed way of using the tool would be the semantic model.

For the official, supported, answer I will let somebody else come and post one. There are plenty of APIs able to do almost everything (well, everything you can do by hand can be scripted), but most of them are technically not documented for public usage...