Few of the Delta tables are not accessible on Lakehouse SQL Endpoint

[aparasar2]

We're encountering an issue with certain Delta tables created through Notebooks in Fabric. While these tables are successfully generated and visible in the Lakehouse, a subset of them are not accessible via the SQL Endpoint. For these problematic tables, the SQL Endpoint displays a perpetual loading icon, and we're unable to execute queries against them. furthermore, we can query these same tables without issue when using Notebooks.

[frithjof_v]

I have seen other users report that if they enable column name mapping on a table, then the table won't appear in SQL Analytics Endpoint.

https://community.fabric.microsoft.com/t5/General-Discussion/Name-based-column-mappings/m-p/4069839

Please vote for this idea:

https://ideas.fabric.microsoft.com/ideas/idea/?ideaid=a93f4fa7-8d03-ee11-a81c-000d3ae52c8f

Another known limitation is that in Lakehouse with schema (preview) enabled, the SQL Analytics Endpoint doesn't work.

[frithjof_v]

It is possible to create Lakehouse tables in Notebook and they will be visible in the SQL Analytics Endpoint. Also when using Spark code. This is normal.

Could you explain more about the code you used to create the tables? Could you share screenshot? Remember to not share any sensitive information.

Have you made any changes to the tables after creating them (like renaming columns, adding columns, removing columns, etc.)?

[v-nuoc-msft]

Hi @aparasar2 

For your problem, I have found the following related content for you:

Solved: Delta Tables in LakeHouse missing in SQL Endpoint - Microsoft Fabric Community

There are several reasons for summarizing. For example,

The visibility of a table in an SQL endpoint is more likely to be related to how the table was created. If the tables were created using Spark code, they will not be visible in the SQL endpoints.

It could be due to an architecture mismatch. You may need to update the schema of the Delta table or fix any schema mismatches.

The increment table may not have been created successfully.

You can try these methods to solve the problem.

Regards,

Nono Chen

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[smpa01]

@v-nuoc-msft can you please elaborate on 'If the tables were created using Spark code, they will not be visible in the SQL endpoints.'

I am intested to know what spark code can create tables that would be invisible to sql endpoint. I am currently building managed delta tables for a number of data tables and dont want to make a mistake now to pay heavily later.

All my tables are created using Delta table builder API.