Fabric shortcut - is it possible in warehouse?

[anandfarmers]

Hi All,

Shortcuts is possible in Lakehouse (and KQL - I have not tried this).

But is NOT possible to create a shortcut in Warehouse.

So if a DIM is required across multiple warehouses, then I need to load it to each warehouse resulting in data duplication and data consistency issues.

Question: How is the single copy of data applicable in the above scenario?

I can create a view in one warehouse based on a table in another warehouse. This can avoid the above problem of duplicate DIMs.

Or link the whole warehouse A to the warehouse B - which is not ideal as the whole of warehouse A tables will be exposed to warehouse B. Also it is not persisted - when you exit the warehouse view, the link is dropped. 

Not having the ability to create a shortcut in Warehouse is a major drawback in Shortcut capability.

Any thoughts how this can be hadled using shortcuts?

[frithjof_v]

Clone tables will not reflect future changes to the data in the source table, right?

So will get out-of-sync with the data in the source table, once the data in the source table gets updated.

So they are different from shortcuts, which always reflect the up-to-date data in the source table (shortcut target).

(unless I misunderstood something ☺️)

Here is link to idea: https://ideas.fabric.microsoft.com/ideas/idea/?ideaid=e5e4905c-7c7a-ee11-a81c-000d3adc66b1

[anandfarmers]

Hi @frithjof_v ,

Yes. you are right.

If using table clone that needs to be part of the ETL logic to create it every day (or whatever frequency). But since it is only a meta data operation, it still maintains one copy of data.

Alternative is creating a View - which will be a one time operation. It will be in synch all the time with the base table(s).

But ideal solution in Shortcut. It was available in Nov 2023 but disapperaed recently. It is a step backward for Fabric in my opinion.

If the concept of "one copy of data" is true, then Shortcut should be consistent across all Fabric items.

Here is MS document link (Nov 2023):

Create OneLake shortcuts - Microsoft Fabric | Microsoft Learn

[frithjof_v]

@anandfarmers Agree ☺️

SQL Views cannot be queried in Direct Lake mode, it will fall back to DirectQuery.

I am thinking a clone will probably be removed from semantic model relationships if the clone is deleted (and recreated). So we would need to recreate the relationship manually each time clone gets deleted and recreated (if using it for Power BI). PS. I haven't tested this, maybe relationships would remain even if deleting and recreate clone, I don't know ☺️

I would prefer shortcuts ☺️

[anandfarmers]

@frithjof_v ,

Thanks for poinitng out the Direct Lake mode. As you mentioned the View does not connect via Direct Lake and fallsback to Direct Query. 

In the MS document it states:

---

---

Before using Direct Lake, you must provision a Lakehouse with one or more delta tables in a workspace hosted on a supported Power BI or Microsoft Fabric capacity.

Learn about Direct Lake in Power BI and Microsoft Fabric - Power BI | Microsoft Learn

So confusing !

[anandfarmers]

Hi @v-nikhilan-msft ,

Would it be possible to get the technical feedback from MS regarding this issue.

As @frithjof_v pointed out, presently there is no solution for cross warehouse DIMs other than loading the DIM table data into each warehouse. The whole "One copy data" is no more?

[v-nikhilan-msft · Accepted solution]

Hi @anandfarmers 
Thanks for using Fabric Community.
Currently Shortcuts can be created only in Lakehouse but not in Warehouse.
To create duplicate tables within same warehouse you can use Clone Table option. You can refer to this document for more information: Link1

Table clones address a different data management when compared to shortcuts. While shortcuts focus on linking to external data sources within a single warehouse, clones create isolated copies of internal tables for specific purposes. 

Appreciate if you could share the feedback on our feedback channel ,which would be open for the user community to upvote & comment on. This allows our product teams to effectively prioritize your request against our existing feature backlog and gives insight into the potential impact of implementing the suggested feature.

Hope this helps. Please let me know if you have any further queries.
 

[anandfarmers]

HI @v-nikhilan-msft ,

Thank you for the detail reply.

I have tried Clone table function and it works well.

I will check the Fabric Ideas site and vote if there is any related ideas.

Thanks,

Anand