Why is default semantic modelling experience inside SQL analytics endpoint/Warehouse?

[fabricator1]

For the default semantic model, why is the 'Open data model' option greyed out in the semantic model item, and instead the modeling experience for the default semantic model is inside the Data Warehouse or Lakehouse SQL Analytics Endpoint?

Do the relationships we set up here, have any implications for the Lakehouse / Data Warehouse?

Curious to learn more about this. Thanks! 😀

Default semantic model, 'Open data model' greyed out:

SQL Analytics Endpoint / Data Warehouse. Why do we do modelling here, instead of inside the Default semantic model item?

[fabricator1]

Hi @v-gchenna-msft 

I have created a new post as you suggested 😀
Do model relationships impact SQL queries on SQL A... - Microsoft Fabric Community

Thank you!

Currently I don't have suggestions regarding the default semantic model.
I'm just trying to understand how it works and how it relates to the SQL Analytics Endpoint and the Data Warehouse.

[v-gchenna-msft · Accepted solution]

Hi @fabricator1 ,

Reply from Internal Team -
" Currently this is as per design, Semantic Model is created by default to allow PBI reports to leverage the data natively without manually creating a semantic model. If you are looking at table structures, you should open the visual editor/ model when you are in the DW or LH SQL Analytics endpoint.

The default semantic model is a direct reflection of the physical structure. If you want to modify the physical structure open the DW or LH artifact and switch to Model.

There are multiple requests in this area - including not showing the default semantic model as well as not refreshing it (which is an option available today) "

We haven't recieved any thing more information from internal. 

Thank you.

[v-gchenna-msft]

Hi @fabricator1 ,

Thanks for using the Fabric Community.
At this time, we are reaching out to the internal team to get some help on this .
We will update you once we hear back from them.

[v-gchenna-msft]

Hi @fabricator1 ,

Apologies for the delay in reply from our side.

Semantic Model is created by default to allow PBI reports to leverage the data natively without manually creating a semantic model. If you are looking at table structures, you should open the visual editor/ model when you are in the DW or LH SQL Analytics endpoint.

At this time, the relationships defined do not have any impact on the physical DW/LH. Team is curious on what your expectations are in this area. If you can share more around what you would like to see?

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.

[fabricator1]

Thank you @v-gchenna-msft !

Currently I am just trying to understand. 

I also found this thread Solved: Re: Building relationship in Fabric Datawarehouse ... - Microsoft Fabric Community and this documentation article Define relationships in data models for data warehousing - Microsoft Fabric | Microsoft Learn which says that when I am creating relationships in the model view, it is automatically being created PK and FK constraints in the SQL tables in order to increase query speed.

I read in this documentation article Primary, foreign, and unique keys - Microsoft Fabric | Microsoft Learn that the PK and FK constraints are not enforced. But the article also says that the (not enforced) primary key constraint may lead to error in query results if there are duplicate values in the column which has the primary key (not enforced) constraint. 

So it seems that adding relationships may lead to some errors when running SQL queries on the SQL Analytics Endpoint / Data Warehouse - if the data is not in accordance with the assumptions made in the relationships. Primary, foreign, and unique keys - Microsoft Fabric | Microsoft Learn

It seems there are no checks in the Lakehouse or Warehouse that will prevent duplicate values in the primary key column, because the constraints are not enforced.

I'm just trying to understand 😀 Thank you!

[v-gchenna-msft]

Hi @fabricator1 ,

Thanks for using Fabric Community and posting your query.

Can you please create a new post as the initial ask is different from your issue?

We will definitely look into the issue and help.

FYI: There are multiple requests coming up in this area - including not showing the default semantic model as well as not refreshing it (which is an option available today).
If you got a chance to provide a feedback, share more around what you would like to see at that area? please share the link here so it helps in internal tracking.

Thanks for understanding.