Forum Discussion
Report return error in visual after promotion in deployment pipeline, that is not seen in 1st stage
- 1 year ago
Yes, we have had the issue resolved with Microsoft with the initiation of a manual refresh of the semantic model. This is even though the continuous sync is turned on for automatic updates through direct lake.
The explanation is that calculations do not automatically sync in the new workspace after deployment, so a manual refresh is necessary for them to display correctly. We're adding this to our deployment procedure to ensure this is done going forward.
The 1,000,000 row error in QA happens because the promoted semantic model isn’t identical to Test — field parameters, relationships, or RLS are being applied differently, so queries return bigger resultsets.
Fix:
compare Test vs QA model in Tabular Editor/XMLA, check field parameters, RLS, and Lakehouse shortcuts. Re-deploy model if mismatched.