Lakehouse T-SQL Endpoint syncing issues (force update fix)

One of the reoccurring issues that people face in Fabric (and post about here!) is the delay it can take for the T-SQL endpoint of a Lakehouse to update, when you make changes to Lakehouse data and tables (via the Lakehouse/ Spark endpoint).

This is most problematic when you have a multi-step ETL/ELT process that first updates some Lakehouse table, then reads that data (via the T-SQL endpoint), using a T-SQL script in a Warehouse.

Well, there is a workaround to force the update of the T-SQL endpoint, using Semantic Link. This workaround was provided initially by the Microsoft product manager for the Lakehouse T-SQL endpoint.

Some links on this:

Hopefully they improve the native functionality in the future, so that forced updates are not required, but in the short-term, this should help a lot of people.

4 comments