Forum Discussion
Differences between Data Warehouse and Lakehouse
Anonymous DennesTorres Hi, Dennes.. now May 2024.. do you have more clarity about your original questions? can you share this here pls.. after read all post.. It is still not clear if there is a different storage between DWH and LH and the impact in our designs
Hi,
During the evolution happening these months, my conclusion is that the Data Warehouse is like a PaaS version of a lakehouse.
The Data Warehouse has features to auto-manage the underlying onelake storage. It has some internal structures we don't have access to and this auto-manage of the onelake storage, which we shouldn't try to handle directly.
The lakehouse, on the other hand, makes the onelake storage an open book to us and some management tasks becomes our responsibility.
For example, in the lakehouse, we can make time-travel as we wish and the maintenance to improve the performance - and prevent time-travel - is up to us. In the Data Warehouse, the time-travel is already limited to 7 days, because the underlying maintenance is automatically applied by the data warehouse.
You can check the time-travel details here: https://www.youtube.com/watch?v=ZVgizgPQRS8&list=PLNbt9tnNIlQ5TB-itSbSdYd55-2F1iuMK&index=9
In the end, a lakehouse is great for a Bronze layer because the Files area, a Data Warehouse is great for the Gold layer because the support to multiple schemas and RLS. The rest, depends a lot.
Kind Regards,
Dennes