Forum Discussion
Advocate IAutomating Delta Table Updates from folders in Lakehouse
Hi,
I suppose the shortcuts you create from the ADLS storage are landed in the Files section of your Lakehouse. Those files are always up-to-date with the source (ADLS storage). To get an updated Delta table, you should use one of the options to copy data from the Files section to the Tables. This can be done with copy data in a data pipeline, DataFlow Gen2 or a notebook where you write your own (Python) script.
Additionally, you could use Data Activator to automatically run the pipeline or notebook once a new file in ADLS is landed. Trigger Fabric items - Microsoft Fabric | Microsoft Learn, but it depends how your data is changed in the ADLS if the trigger exists.
Advocate II don't want to move data from ADLS to OneLake. Is there any option to access the data without moving it from one place to another?
- FabianSchut
Solution Sage
If you want to create a shortcut directly from the ADLS to the Tables section in your Lakehouse, you should have your data in delta format in ADLS. If your source data is not in delta format, you should convert it. You can still use a pipeline for that, but then the source and destination are both ADLS and you can create a shortcut in the Lakehouse from the delta files.
frithjof_vCommunity Champion
Another option is to create an external (unmanaged table) in the Table section in the Fabric Lakehouse, based on the file shortcut in the File section. The external table will always be updated, without the need to copy data.
However, this external table won't work with the SQL Analytics Endpoint or Direct Lake afaik.
I think the options provided by FabianSchut are generally your best options in this case.
If your data in ADLS is in Delta Table format, then you can create the shortcut directly in the Table section of the Lakehouse, without copying data.
Please note: if you will use Direct Lake with the table, there is an advantage of creating the Delta Table using Fabric (notebooks, data pipeline, Dataflow Gen2). That is that the Delta Table will be V-Ordered when you create it using Fabric. V-Ordering is a performance booster for Direct Lake.
Could you share what will be your use cases for this particular data inside Fabric?
Will you use it for Power BI (Direct Lake or Import mode), T-SQL queries, data science/ML, building a data model, etc.?