Historical Reporting

[Mswedorske]

How do you handle reporting monthly trends within a data lakehouse?  Can this be done with timetravel to get the table state at the end of each month or is it better practice to build a data warehouse with SCD types?  

We are new to databricks and lakehouses.  

[daniel_sahal · Accepted solution]

@Mswedorske IMO it would be better to use SCD.
When you do VACUUM on a table, it removes the data files that are necessary for Time Travel, so it's not a best choice to rely on Time Travel.